Skip to main content
NCBI home page
JAMA Network logoLink to JAMA Network
. 2026 Jul 10;7(7):e262108. doi: 10.1001/jamahealthforum.2026.2108

Eco-Focused Menu Labels on Full Meal Orders From Fast-Food Restaurants

A Randomized Clinical Trial

Julia A Wolfson 1,2,✉, Anna Claire Tucker 1, Alexandria E Reimold 3, Aviva Musicus 4,9, Nina Carr 1, Brent F Kim 5,6, Raychel Santo 7, Clara Cho 7, Daphne Abigail Barquera Guevara 1, Erin Choi 1, Cindy W Leung 4, Christina A Roberto 8, Jennifer Falbe 3
PMCID: PMC13355149  PMID: 42430158

Key Points

Question

What effects do ecolabel designs have on the healthfulness and environmental sustainability of full meal orders at 2 fast-food restaurants in a nationally representative sample?

Findings

In this randomized clinical trial with 6210 adults in the US, the traffic-light label outperformed other label designs and the control label. The traffic-light label led to meal selections with 30% lower greenhouse gas emissions from a sandwich restaurant and 15% lower greenhouse gas emissions and slightly more healthful meals ordered from a burger restaurant compared to the control label.

Meaning

Traffic-light–style ecolabels may be an effective strategy to promote more sustainable choices from fast-food restaurants.

Abstract

Importance

Promoting environmentally sustainable dietary patterns can benefit both population and planetary health. Ecolabels on restaurant menus could promote healthful and sustainable choices, but evidence is needed on effects in different restaurant types.

Objective

To test the effects of different ecolabel designs on menus from 2 fast-food restaurants on the healthfulness of full meal orders, compared to a control label.

Design, Setting, and Participants

This randomized clinical trial was conducted online in a nationally representative sample of adults (≥18 years) in the US in April 2025. Participants were recruited from the NORC AmeriSpeak survey panel to participate in this experiment.

Intervention

Participants were randomized via an automated process within the survey platform to 1 of 5 label conditions: (1) control: QR code; (2) low–climate-impact labels; (3) traffic-light labels (TLLs); (4) high–climate-impact labels; or (5) grade-scale labels and asked to order lunch from 2 fast-food menus (1 burger-focused and 1 sandwich-focused).

Main Outcomes and Measures

The primary outcome was the healthfulness of meals ordered based on a modified Nutrient Profile Index (NPI) score. Other outcomes included the estimated greenhouse gas emissions (GHGE), calorie and nutrient content, and cost of selected meals, as well as perceptions of labels and knowledge about the climate contribution of different foods.

Results

Among the 6210 participants (3274 [51.0%] women; mean [SD] age, 48.1 [18.1] years), 1236 participants were randomized to the control group; 1249 to low–climate-impact labels, 1243 to high–climate-impact labels, 1235 to TLLs; and 1247 to grade-scale labels. Demographic characteristics were similar across groups. Only the TLL was effective at promoting more healthful food choices in the burger restaurant compared to the control label (mean [standard error] NPI score: 49.93 [0.18] points vs 49.25 [0.18] points; P < .001). There were no ecolabel effects on NPI score in the sandwich restaurant. Compared to control, the TLL led to 15.2% lower GHGE for meals in the burger restaurant and 29.8% lower GHGE for meals in the sandwich restaurant. The TLL was the most noticeable (78.30% [95% CI, 75.35%-81.26%]; P < .001) and had the highest percentage of participants who correctly reported what the label was about (88.72% [95% CI, 86.04%-91.39%]; P < .001). There were no differences in the overall cost of the meals selected by label condition.

Conclusions and Relevance

Findings from this web-based randomized clinical trial showed that TLL-style ecolabels on restaurant menus may be an effective strategy to promote more sustainable and, in some instances, more healthful food choices without increasing consumer costs.

Trial Registration

ClinicalTrials.gov Identifier: NCT06909019

This randomized clinical trial tests the effects of different ecolabel designs on menus from 2 fast-food restaurants on the healthfulness of full meal orders, compared to a control label.

Introduction

The food system,1,2 especially red meat production,2,3 is a large contributor to global greenhouse gas emissions (GHGE) and climate change. Red and processed meat consumption is also associated with numerous adverse health outcomes, with stronger evidence for processed meat.4,5,6,7,8 In contrast, more sustainable diets are associated with better diet quality.2,9,10,11,12 Therefore, promoting more environmentally sustainable dietary patterns is a key strategy that can benefit both population and planetary health.

On a typical day, one-third of adults in the US eat fast food, a major source of red meat in the US diet.13,14,15 Therefore, even small shifts toward more sustainable menu offerings and consumer choices in fast-food restaurants could lead to meaningful improvements in diet quality and human and planetary health. Sustainability is a trend in the restaurant sector,16,17,18 and numerous food brands have begun voluntarily implementing ecolabeling schemes.18,19,20 It is important to understand how different ecolabel designs affect both the healthfulness and sustainability of consumer food choices, particularly in fast-food restaurants, where most menu items are not healthful.21,22,23

Ecolabels may promote more healthful and/or more sustainable food selections.24,25,26,27,28,29,30 However, in restaurant settings, the few ecolabel studies to date have not assessed full meal orders,26 have not compared label effects across different restaurant types,25,26,30 and only one (using a nonrepresentative sample) has directly compared different label designs30 (eg, warning labels, grade-scale labels [eg, A-F] or traffic-light labels [TLL; ie, red, yellow, or green]).31,32,33 Ecolabels may increase perceptions of the healthfulness of items labeled as sustainable,26,34,35 whether that perception is justified or not. Further, perceptions of labeled items and efficacy of labels may differ depending on menu composition and restaurant type.

This study’s objective was to test the effect of different ecolabels on the healthfulness of full meal orders in 2 fast-food restaurants, modeled after popular burger- and sandwich-focused chains. Secondary objectives were to assess label effects on GHGE, calories, and nutrients of selected meals, and understanding and perceptions of the ecolabels and labeled menu items. Given existing evidence,31,32,33 we hypothesized that, relative to a control label, the TLL would lead to the most healthful food choices, followed by the high-impact label, then the grade-scale label.

Methods

Study Design and Participants

This randomized clinical trial followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline. This trial was approved by the Johns Hopkins Bloomberg School of Public Health Institutional Review Board. Written informed consent was obtained from all participants. The study was conducted from April 1 to April 23, 2025. Analyses were preregistered on AsPredicted.com (219376).

Participants were recruited from the National Opinion Research Center (NORC) AmeriSpeak panel, a probability-based, nationally representative panel of US households recruited by telephone, mail, and field interviews.36 Adults in the US (aged ≥18 years) were eligible for this study. The target sample size was 6000 based on a priori power calculations to ensure a small detectable effect (Cohen d = 0.14) with greater than 80% power for the primary outcome. The trial protocol can be found in Supplement 1.

Procedures

Participants were invited to complete a survey for a research study “to understand things you consider when ordering food from a fast-food restaurant.” First, participants were instructed to order lunch from each restaurant. Participants were randomized 1:1:1:1:1 to view menus with one of the following label conditions: (1) control: QR code labels on all items; (2) low–climate-impact labels on select items; (3) TLLs (red, yellow, or green) on all items; (4) high–climate-impact labels on select items; or (5) grade-scale labels (A-F) on all items (eAppendix 4 in Supplement 2). Two menus based on popular burger- and sandwich-based fast-food restaurant chains were presented to participants in random order and were labeled according to participants’ assigned label condition. Menus also displayed prices (extracted and averaged from the restaurants’ websites in 16 US cities (eAppendix 1 in Supplement 2), calorie information (per US regulations), and explanatory text at the top describing the meaning of the labels. Participants were required to choose 1 main item and could separately select up to 3 additional items from a section displaying sides, beverages, and desserts. To encourage realistic choices, we used a minor deception: participants were informed that 1 in 25 respondents would receive a coupon for the meals they select, in addition to their survey-completion compensation. Participants were compensated in AmeriSpeak points, which can be redeemed for prizes. Those who “won” the raffle were compensated with additional points equal to the value of their selected meals. eTable 1 in Supplement 2 lists all items on the 2 menus. After ordering, participants completed a survey.

Measures

Healthfulness of the Selected Meal

The primary outcome was the mean Nutrition Profile Index (NPI) score of the selected meal. NPI was selected because it has been widely used to measure healthfulness of fast-food restaurant menu items,37,38,39,40,41 and can be calculated using publicly available nutrition information provided by restaurants. NPI is based on the United Kingdom Ofcom Nutrient Profiling Model, which assigns individual foods a score from 0 to 100, with 64 or higher considered healthful.39 Points are awarded for positive components (ie, fiber, protein, fruit, vegetables, and nuts) and subtracted for negative components (ie, calories, saturated fat, sodium, and sugar). Further details on calculating NPI score are available elsewhere.39,42 We obtained nutritional and serving size information from the restaurant websites (eAppendix 1 in Supplement 2). Because NPI was originally developed to score individual foods, we used a modified NPI score to score full meals, calculating the mean NPI score of all foods in the meal weighted by the mass of each food. Since NPI scoring does not differentiate beverages well, beverages were excluded from mean NPI meal scores and assessed separately as a secondary outcome. Secondary outcomes were preregistered and are described below. eTable 2 in Supplement 2 describes item wording and response options.

GHGE of the Selected Meal

GHGE, kilograms carbon dioxide equivalent (kgCO2e) per meal, were calculated by multiplying the estimated quantity of each ingredient in a menu item by an associated GHGE factor from the Coolfood Pledge and Coolfood Meals Calculators,43,44 then summing the result over all ingredients in all menu items in each meal. GHGE factors included supply chain emissions and carbon opportunity costs. Beverages were excluded from meal-level GHGE due to limitations in accurately assessing their GHGE. Following methods described by Waite and Blondin,45 thresholds for labels (eg, yellow vs green TLLs) were set relative to average meal-related GHGE, and the GHGE reductions needed to align with climate mitigation goals. Further description of GHGE methods and associated thresholds are available in eAppendix 2 in Supplement 2. We also assessed whether participants selected a sustainable (ie, assigned green TLL) main entrée item (eg, burger, sandwich).

Perceptions of Labeled Items

We showed participants 4 menu items (labeled according to condition) from each restaurant representing 4 permutations of high/low healthfulness and high/low sustainability and asked about perceptions of appeal, healthfulness, and the climate impact of each item. The order of the restaurant and items within each restaurant were randomized. Responses were measured on 7-point Likert scales.

Perceived Message Effectiveness

Participants viewed the label corresponding to their assigned condition and answered questions about perceived message effectiveness using a modified version of the University of North Carolina Perceived Message Effectiveness (PME) scale.46,47 The modified PME included 4 items, and scores ranged from 1 to 5 for each item, with higher scores indicating greater message effectiveness. Scores were averaged across the 4 items to create a mean PME scale. We further assessed label believability and the extent to which it grabs attention. All responses were measured on a 5-item unipolar scale.

Noticeability and Label Use

Participants were asked if they “notice any label (other than calories).” Those who reported noticing the ecolabel were asked what the label was about and whether they used the label when deciding what to order.

Knowledge of Climate Impact

Participants were asked to rank the climate impact of 3 foods (tofu, beef, and chicken) from smallest to largest climate impact. We also asked participants to determine which of 2 items from each restaurant (beef burger vs chicken sandwich; ham sandwich vs roast beef sandwich) had the larger climate impact or whether it was the same. In each question, participants were shown images of the menu items in random order and labeled according to condition. The outcome was the percentage of participants responding correctly to both questions.

Total Nutrients of Selected Meal

We assessed the total energy (calories) and nutrient content (saturated fat [g], protein [g], fiber [g], total sugar [g], and sodium [mg]) of selected meals (including beverages). Additional outcomes included NPI score of the main item selected, red meat selection (yes/no), sugar-sweetened beverage selection (yes/no), and total cost of selected meal, including beverages.

Sociodemographic Characteristics

Self-reported sociodemographic characteristics were part of participants’ profiles and were provided by AmeriSpeak. Race categories included Hispanic, non-Hispanic Black, non-Hispanic White, non-Hispanic Other, non-Hispanic multiracial, and non-Hispanic Asian/Pacific Islander. Due to small sample sizes in some categories, a category for other was created to include non-Hispanic with other race, multiracial, and Asian/Pacific Islander.

Statistical Analysis

We used AmeriSpeak’s study-specific survey weights to obtain nationally representative estimates. For primary analyses, we used linear regression models with label condition as the independent variable to estimate the predicted mean NPI score with standard error (SE), stratified by restaurant. We included an interaction term between label condition and whether the burger menu or sandwich menu was presented first in models with a significant interaction. To examine the predicted mean total kgCO2e per meal stratified by restaurant, we used generalized linear models with a gamma family and log link, as GHGE were positively skewed. We tested primary models for interaction by sociodemographic characteristics (sex, age, race and ethnicity, education, income, region, metropolitan/nonmetropolitan residence, and political beliefs), and noticing a label. Linear models with experimental condition as the independent variable were also used to estimate PME, attitudes, beliefs, mean cost, energy, and nutrients of selected meals, stratified by restaurant. We used logistic regression, stratified by restaurant, to estimate the following probabilities by experimental condition: noticing a label, correctly identifying the label, using the label, correctly identifying menu items with higher climate impact, correctly ranking foods from lowest to highest climate impact, selecting a sustainable item, selecting an item containing red meat, and selecting a sugar-sweetened beverage. We used postestimation margins commands to obtain predicted means and probabilities. All analyses were conducted in Stata 18 statistical software (StataCorp LLC). We considered statistical significance at P < .05 with Bonferroni-Holm correction for pairwise comparisons of the 4 label conditions for main analyses including NPI score and GHGE, and within family of outcomes for secondary outcomes. All tests were 2-sided.

Results

Of the 6486 participants who started the survey, 15 participants were excluded due to issues with block order randomization (not preregistered). Additional preregistered exclusions included completing the survey in less than 33% of the median survey duration (n = 159), missing greater than 50% of eligible questions (n = 18), and missing responses after the menu ordering tasks (n = 84) (Figure 1).

Figure 1. CONSORT Flow Diagram.

Flow diagram of participant exclusions, randomization, and analyzed groups. Top center rectangle labeled 6486 Participants clicked on survey. A vertical connector line descends to a right-side light gray rectangle labeled 192 Excluded prior to randomization, followed by three indented lines: 159 Completed survey in less than 22 percent median duration; 18 Missing greater than 50 percent of eligible questions; 15 Experienced issue with block order randomization. The main vertical line continues downward to a central horizontal oval labeled 6294 Randomized into label conditions. From the oval, five diagonal connector lines extend to five light gray rectangles along the lower portion of the figure. From left to right, the first rectangle reads 1236 Allocated to control group, with parenthetical Q R code, and analyzed. The second rectangle reads 1249 Allocated to low-climate impact labels and analyzed. The third rectangle reads 1243 Allocated to high-climate impact labels. The fourth rectangle reads 1235 Allocated to traffic-light labels and analyzed, with a small superscript letter a after analyzed. The fifth rectangle reads 1247 Allocated to grade-scale labels and analyzed, with a small superscript letter b after analyzed. Near the center-right, a smaller light gray rectangle connected by a short arrow from the main vertical pathway reads 84 Excluded for missing questions after ordering task. At the bottom, a wide horizontal light gray bar spans the figure and reads 6210 Analyzed, with downward arrows from the group allocation boxes pointing toward this bar.

Open in a new tab

The final analytic sample included 6210 participants (3274 [51.0%; 95% CI, 49.4%-52.7%] women and 2936 [49.0%; 95% CI, 47.3%-50.6%] men; mean [SD] age 48.1 [18.1] years; 783 [12.0%; 95% CI, 11.0%-13.1%] Black, 1128 [17.8%; 95% CI, 16.5%-19.2%] Hispanic, 3854 [61.0%; 95% CI, 59.3%-62.6%] White, and 445 [9.2%; 95% CI, 8.2%-10.3%] had another race) (Table 1). Frequency of menu item selections overall and across label conditions is presented in eTable 3 in Supplement 2. In the burger-based restaurant, across experimental arms there were differences in frequency of selecting the plant-based burger with the lowest frequency in the control condition (3.4%; 95% CI, 2.4%-4.7%; P < .001) and the highest frequency in the TLL condition (9.8%; 95% CI, 7.7%-12.5%; P < .001). In the burger-based restaurant, individuals in the TLL condition selected more healthful meals than those in the control (mean [SE] NPI score, 49.93 [0.18] vs 49.25 [0.18], respectively; P < .001), and the high-impact (mean [SE] NPI score, 49.45 [0.20]; P = .005) and grade-scale (mean [SE] NPI score, 49.50 [0.20]; P = .002) conditions (Figure 2). There were no significant differences in selections of individual menu items (eTable 3 in Supplement 2) or in NPI scores (Figure 2) by label condition in the sandwich-based restaurant. There were no significant differences in mean calories or nutrients by label condition in either restaurant (eTable 4 in Supplement 2). NPI scores for the entrée selected are reported in eFigure 2 in Supplement 2. There were no significant interactions between label conditions and sociodemographic characteristics in NPI scores in either restaurant.

Table 1. Participant Characteristics Overall and by Experimental Condition.

Characteristic Label, No. (weighted %) [95% CI]
Overall Control Low climate impact Traffic light High climate impact Grade scale
Participants 6210 (100) 1236 (19.9) 1249 (20.1) 1235 (19.9) 1243 (20.0) 1247 (20.1)
Sex
Female 3274 (51.0) [49.4-52.7] 658 (52.2) [48.6-55.9] 663 (50.8) [47.1-54.5] 671 (53.1) [49.4-56.7] 629 (48.6) [44.9-52.3] 653 (50.5) [46.7-54.2]
Male 2936 (49.0) [47.3-50.6] 578 (47.8) [44.1-51.4] 586 (49.2) [45.5-52.9] 564 (46.9) [43.3-50.6] 614 (51.4) [47.7-55.1] 594 (49.5) [45.8-53.3]
Age group, y
18-29 961 (19.6) [18.2-21.1] 191 (19.1) [16.2-22.4] 181 (18.6) [15.6-22.1] 182 (19.2) [16.1-22.7] 214 (21.3) [18.2-24.7] 193 (19.8) [16.6-23.4]
30-44 1751 (25.7) [24.4-27.1] 337 (25.5) [22.5-28.9] 349 (24.1) [21.3-27.2] 368 (27.5) [24.5-30.7] 338 (24.4) [21.4-27.6] 359 (27.1) [24.1-30.4]
45-59 1437 (23.8) [22.4-25.2] 280 (23.1) [20.3-26.2] 295 (24.9) [21.8-28.1] 295 (23.4) [20.5-26.5] 272 (23.8) [20.7-27.3] 295 (23.7) [20.7-27.0]
≥60 2061 (30.9) [29.4-32.4] 428 (32.2) [28.9-35.7] 424 (32.4) [29.1-35.8] 390 (29.9) [26.7-33.3] 419 (30.5) [27.3-33.9] 400 (29.4) [26.2-32.7]
Race and ethnicity
Hispanic 1128 (17.8) [16.5-19.2] 232 (19.6) [16.8-22.8] 216 (17.5) [14.8-20.6] 216 (18.3) [15.5-21.5] 222 (16.0) [13.5-19.0] 242 (17.6) [14.8-20.8]
Non-Hispanic Black 783 (12.0) [11.0-13.1] 159 (12.8) [10.6-15.4] 170 (13.0) [10.7-15.6] 147 (11.9) [9.7-14.6] 144 (10.3) [8.4-12.6] 163 (12.1) [9.9-14.5]
Non-Hispanic White 3854 (61.0) [59.3-62.6] 770 (60.8) [57.1-64.3] 769 (59.4) [55.7-63.0] 784 (61.6) [57.9-65.1] 791 (64.0) [60.4-67.6] 740 (59.1) [55.3-62.8]
Othera 445 (9.2) [8.2-10.3] 75 (6.8) [5.0-9.0] 94 (10.1) [7.9-12.9] 88 (8.1) [6.2-10.6] 86 (9.6) [7.3-12.5] 102 (11.2) [8.7-14.3]
Education
<High school 358 (8.9) [7.9-10.2] 76 (9.9) [7.6-12.8] 71 (8.3) [6.3-10.9] 77 (9.8) [7.5-12.7] 70 (8.6) [6.4-11.4] 64 (8.3) [5.9-11.5]
High school graduate or equivalent 1139 (28.7) [27.1-30.3] 240 (30.1) [26.6-33.8] 225 (28.6) [25.1-32.5] 218 (27.6) [24.2-31.3] 237 (30.4) [26.8-34.2] 219 (26.7) [23.2-30.4]
Some college or associate’s degree 2418 (26.5) [25.2-27.8] 473 (25.8) [23.1-28.7] 500 (27.8) [25.0-30.9] 482 (26.1) [23.4-28.9] 480 (26.4) [23.6-29.3] 483 (26.3) [23.5-29.2]
Bachelor’s degree 1341 (22.5) [21.2-23.9] 246 (20.5) [17.8-23.6] 264 (21.3) [18.5-24.3] 278 (23.7) [20.8-26.8] 275 (23.3) [20.3-26.6] 278 (23.8) [20.9-27.0]
Postgraduate study or professional degree 954 (13.4) [12.4-14.4] 201 (13.7) [11.7-16.1] 189 (13.9) [11.7-16.4] 180 (12.9) [10.8-15.3] 181 (11.4) [9.5-13.5] 203 (15.0) [12.7-17.6]
Household income, $
<30 000 1306 (21.0) [19.7-22.4] 274 (22.0) [19.1-25.3] 261 (19.6) [16.9-22.7] 264 (21.4) [18.5-24.6] 260 (22.1) [19.1-25.5] 247 (20.0) [17.1-23.2]
30 000 to <60 000 1508 (23.9) [22.5-25.3] 307 (25.9) [22.9-29.2] 320 (25.6) [22.5-28.8] 299 (23.7) [20.7-26.9] 298 (23.3) [20.4-26.6] 284 (20.8) [18.1-23.9]
60 000 to <100 000 1548 (24.6) [23.2-26.0] 319 (25.0) [22.0-28.3] 312 (24.8) [21.7-28.0] 307 (24.9) [21.9-28.2] 292 (22.6) [19.7-25.8] 318 (25.7) [22.6-29.1]
≥100 000 1848 (30.5) [29.0-32.1] 336 (27.0) [23.9-30.3] 356 (30.1) [26.8-33.6] 365 (30.0) [26.8-33.4] 393 (31.9) [28.5-35.5] 398 (33.5) [30.0-37.2]
Region
Northeast 827 (17.3) [16.0-18.6] 155 (16.3) [13.7-19.3] 172 (18.4) [15.6-21.6] 160 (16.6) [13.9-19.7] 171 (17.6) [14.8-20.7] 169 (17.5) [14.7-20.7]
Midwest 1691 (20.5) [19.3-21.7] 347 (21.7) [19.1-24.6] 361 (21.4) [18.7-24.3] 341 (20.6) [18.0-23.4] 316 (18.7) [16.3-21.4] 326 (20.0) [17.5-22.9]
South 2099 (38.4) [36.8-40.1] 438 (40.9) [37.4-44.6] 397 (37.0) [33.5-40.6] 417 (38.9) [35.4-42.6] 443 (40.3) [36.6-44.0] 404 (35.2) [31.7-38.8]
West 1593 (23.6) [22.4-25.3] 296 (21.1) [18.2-24.2] 319 (23.3) [20.3-26.5] 317 (23.9) [21.0-27.0] 313 (23.4) [20.4-26.7] 348 (27.3) [24.0-30.9]
Metropolitan residence
Nonmetropolitan 879 (13.5) [12.4-14.6] 195 (15.5) [13.1-18.2] 179 (13.1) [10.9-15.7] 183 (14.0) [11.8-16.6] 159 (12.6) [10.5-15.1] 163 (12.2) [10.1-14.7]
Metropolitan 5331 (86.5) [85.4-87.6] 1041 (84.5) [81.8-86.9] 1070 (86.9) [84.3-89.1] 1052 (86.0) [83.4-88.2] 1084 (87.4) [84.9-89.5] 1084 (87.8) [85.3-89.9]
Political beliefs
Unknown 43 (0.7) [0.5-1.1] 5 (0.5) [0.2-1.3] 6 (0.5) [0.2-1.7] 12 (1.1) [0.5-2.5] 9 (0.7) [0.3-1.6] 11 (0.9) [0.4-1.7]
Very liberal 817 (12.1) [11.1-13.2] 181 (14.1) [11.7-16.9] 162 (11.8) [9.8-14.3] 163 (12.4) [10.3-14.8] 156 (10.1) [8.3-12.3] 155 (12.1) [9.9-14.7]
Somewhat liberal 662 (10.4) [9.4-11.5] 137 (10.5) [8.5-12.8] 137 (10.8) [8.7-13.4] 133 (10.2) [8.1-12.7] 128 (9.6) [7.7-11.9] 127 (10.7) [8.3-13.8]
Moderate 3052 (49.8) [48.1-51.4] 588 (49.1) [45.4-52.7] 605 (49.3) [45.7-53.0] 610 (49.1) [45.5-52.8] 623 (52.8) [49.1-56.5] 626 (48.4) [44.7-52.2]
Somewhat conservative 872 (14.7) [13.6-15.9] 175 (14.4) [12.0-17.1] 159 (13.2) [10.9-16.0] 173 (15.0) [12.6-17.8] 170 (13.9) [11.6-16.6] 195 (17.2) [14.5-20.2]
Very conservative 764 (12.3) [11.3-13.4] 150 (11.5) [9.5-13.8] 180 (14.2) [11.9-17.0] 144 (12.2) [10.0-14.8] 157 (12.8) [10.5-15.5] 133 (10.7) [8.6-13.2]
Open in a new tab
a

Due to small sample sizes in some categories, the other group was created to include non-Hispanic with other race, multiracial, and Asian/Pacific Islander.

Figure 2. Bar Chart of Mean Predicted Nutrition Profile Index (NPI) Score and Greenhouse Gas Emissions of Meals by Experimental Condition and Restaurant (N = 6210).

Four-panel bar charts of N P I scores and greenhouse gas emissions by label and restaurant. Four panels arranged in a two by two grid, labeled A, B, C, and D in small boxed letters at each panel’s upper left. Panel A title: Burger restaurant N P I score. Vertical axis labeled Mean predicted N P I score, ranging from 45 to 70 with tick marks at 45, 50, 55, 60, 65, and 70. Horizontal axis labeled Label with five categories: Control, Low climate impact, Traffic light, High climate impact, Grade scale. Five light blue bars with black outlines and black error bars. Approximate bar heights: Control about 49 point 3; Low climate impact about 49 point 5; Traffic light about 50 point 0; High climate impact about 49 point 5; Grade scale about 49 point 6. Above selected pairs, thin horizontal comparison lines with small filled triangles and P values: between Low climate impact and Traffic light, P equals .02; between Traffic light and High climate impact, P less than .001; between Traffic light and Grade scale, P equals .005; between High climate impact and Grade scale, P equals .002. Panel B title: Sandwich restaurant N P I score. Same vertical axis label and range. Same five label categories on the horizontal axis. Five light blue bars with black outlines and black error bars, all near 64: Control about 63 point 9; Low climate impact about 64 point 3; Traffic light about 64 point 4; High climate impact about 64 point 4; Grade scale about 64 point 2. No P value annotations visible in panel B. Panel C title: Burger restaurant greenhouse gas emissions. Vertical axis labeled Greenhouse gas emissions, kg C O 2 e, ranging from 0 to 25 with tick marks at 0, 5, 10, 15, 20, and 25. Same five label categories on the horizontal axis. Five green bars with black outlines and black error bars. Approximate heights: Control about 20 point 8; Low climate impact about 20 point 5; Traffic light about 17 point 7; High climate impact about 21 point 1; Grade scale about 20 point 0. P value annotations above comparisons: between Low climate impact and Traffic light, P equals .004; between Traffic light and High climate impact, P less than .001; between Traffic light and Grade scale, P equals .001. Panel D title: Sandwich restaurant greenhouse gas emissions. Same vertical axis label and range as panel C and same five label categories. Five green bars with black outlines and black error bars. Approximate heights: Control about 19 point 2; Low climate impact about 18 point 2; Traffic light about 13 point 5; High climate impact about 17 point 6; Grade scale about 15 point 8. P value annotations above comparisons: between Low climate impact and Traffic light, P equals .001; between Traffic light and High climate impact, P less than .001; between Traffic light and Grade scale, P equals .004; between High climate impact and Grade scale, P equals .006.

Open in a new tab

Statistical significance was set at P < .05 with Bonferroni-Holm correction for multiple comparisons of 4 label conditions. Meals exclude beverages. A, Postestimation margins after weighted simple linear regression and including an interaction term for the order restaurant menus were presented. B, Postestimation margins after weighted simple linear regression. C and D, Postestimation margins after weighted generalized linear model with a gamma family and log link. The error bars indicate 95% CIs.

In both restaurants, participants in the TLL condition selected meals with lower GHGE than the control condition (burger restaurant: 15.2% lower; mean [SE], 17.9 [0.7] vs 21.1 [0.7] kgCO2e; P = .001; sandwich restaurant: 29.8% lower; mean [SE], 13.6 [0.8] vs 19.4 [1.0] kgCO2e; P < .001) (Figure 2). In both restaurants, the TLL condition also had lower GHGE compared to the low-impact condition (burger restaurant: mean [SE], 17.9 [0.7] vs 20.1 [0.7] kgCO2e; sandwich restaurant: mean [SE], 13.6 [0.8] vs 18.3 [0.9] kgCO2e, respectively) and the high-impact condition (burger restaurant: mean [SE], 17.9 [0.7] vs. 21.4 [0.8] kgCO2e; sandwich restaurant: mean [SE], 13.6 [0.8] vs. 17.6 [1.2] kgCO2e, respectively). Compared to the control, the TLL condition also had higher probability of selecting a sustainable (green-labeled) item in both restaurants (eFigure 3 in Supplement 2). In the burger restaurant, the TLL condition had a lower probability of selecting a red meat item compared to the control, low-impact, and high-impact conditions (eFigure 4 in Supplement 2).

The TLL was the most noticeable (78.30% [95% CI, 75.35%-81.26%]; P < .001) compared to the control (42.85% [95% CI, 39.23%-46.48%]) (eTable 5 in Supplement 2) and had the highest percentage of participants who correctly reported what the label was about (88.72% [95% CI, 86.04%-91.39%]; P < .001). Less than half of participants reported using the ecolabels when ordering; the highest percentage was in the TLL condition (38.9% [95% CI, 34.5%-43.2%]). Differences by label condition in the healthfulness of the meal orders did not differ based on label noticeability (eFigure 1 in Supplement 2).

Table 2 presents perceptions of 4 menu items in each restaurant by label condition. In both restaurants, the TLLs and the grade-scale labels were significantly different than the control label in rankings of the climate impact of all menu items, with lower and higher GHGE items correctly perceived as having lower and higher climate impact, respectively.

Table 2. Mean Predicted Participant Perceptions of Menu Items at a Burger-Focused and Sandwich-Focused Restaurant by Treatment Condition (N = 6210)a.

Menu item Control, mean (95% CI) (n = 1236) Label
Low climate impact (n = 1249) Traffic light (n = 1235) High climate impact (n = 1243) Grade scale (n = 1247)
Mean (95% CI) P value Mean (95% CI) P value Mean (95% CI) P value Mean (95% CI) P value
How appealing do you think this menu item is? b
Burger restaurant
Plant-based burger 4.02 (3.88-4.16) 3.76 (3.62-3.91) .01 3.86 (3.71-4.01) .12 3.80 (3.65-3.94) .03 3.80 (3.65-3.94) .03
8-Piece chicken nuggets 4.40 (4.28-4.52) 4.18 (4.05-4.31) .02 4.35 (4.22-4.48) .63 4.28 (4.15-4.40) .17 4.32 (4.20-4.44) .37
Burger 5.00 (4.88-5.12) 4.79 (4.67-4.91) .02 4.84 (4.72-4.96) .07 4.82 (4.70-4.95) .04 4.82 (4.70-4.94) .04
Bacon burger 4.65 (4.51-4.79) 4.40 (4.26-4.54) .01 4.49 (4.35-4.63) .12 4.45 (4.31-4.60) .06 4.38 (4.23-4.52) .009
Sandwich restaurant
6-in Tuna 3.89 (3.75-4.04) 3.97 (3.83-4.10) .46 4.08 (3.94-4.22) .06 3.88 (3.74-4.02) .93 3.88 (3.74-4.02) .91
Grilled chicken salad 5.09 (4.99-5.20) 4.95 (4.84-5.06) .07 4.99 (4.87-5.11) .20 4.96 (4.84-5.07) .09 4.99 (4.88-5.10) .18
Roast beef salad 4.10 (3.97-4.23) 4.10 (3.97-4.23) .99 3.92 (3.79-4.06) .06 3.95 (3.82-4.07) .09 3.92 (3.79-4.05) .06
12-in Steak and cheese 4.90 (4.78-5.02) 4.83 (4.70-4.97) .46 4.67 (4.54-4.80) .01 4.69 (4.55-4.83) .02 4.71 (4.58-4.85) .04
How healthy do you think this item is? c
Burger restaurant
Plant-based burger 3.53 (3.42-3.65) 3.73 (3.61-3.84) .02 3.85 (3.73-3.97) <.001 3.65 (3.54-3.76) 0.15 3.78 (3.65-3.90) 0.005
8-Piece chicken nuggets 2.87 (2.78-2.97) 2.97 (2.87-3.07) .17 3.07 (2.98-3.17) .004 2.87 (2.77-2.96) 0.92 3.07 (2.97-3.17) 0.005
Burger 2.80 (2.71-2.90) 2.74 (2.65-2.83) .35 2.56 (2.47-2.65) <.001 2.75 (2.66-2.85) 0.48 2.70 (2.60-2.80) 0.15
Bacon burger 2.21 (2.11-2.30) 2.12 (2.03-2.22) .21 2.02 (1.93-2.11) .005 2.16 (2.07-2.26) 0.52 2.06 (1.96-2.16) 0.03
Sandwich restaurant
6-in Tuna 4.46 (4.35-4.56) 4.53 (4.43-4.64) .31 4.63 (4.53-4.73) .02 4.48 (4.38-4.57) 0.79 4.54 (4.45-4.64) 0.24
Grilled chicken salad 5.31 (5.22-5.39) 5.30 (5.21-5.39) .89 5.34 (5.25-5.44) .55 5.27 (5.19-5.36) 0.61 5.26 (5.17-5.35) 0.48
Roast beef salad 4.52 (4.42-4.62) 4.54 (4.43-4.64) .83 4.12 (4.02-4.23) <.001 4.25 (4.15-4.35) <0.001 4.33 (4.22-4.44) 0.01
12-in Steak and cheese 3.21 (3.11-3.31) 3.15 (3.05-3.24) .39 2.93 (2.83-3.03) <.001 3.07 (2.97-3.16) 0.05 3.00 (2.89-3.10) 0.004
What do you think the climate impact of this item is? d
Burger restaurant
Plant-based burger 4.41 (4.31-4.51) 4.60 (4.52-4.69) .003 4.74 (4.66-4.83) <.001 4.48 (4.38-4.57) .35 5.01 (4.89-5.13) <.001
8-Piece chicken nuggets 4.27 (4.17-4.37) 4.44 (4.35-4.53) .02 4.56 (4.48-4.64) <.001 4.38 (4.29-4.46) .13 4.53 (4.44-4.62) <.001
Burger 4.12 (4.02-4.22) 3.95 (3.85-4.04) .02 3.20 (3.11-3.29) <.001 3.41 (3.31-3.51) <.001 3.46 (3.36-3.57) <.001
Bacon burger 3.89 (3.78-4.00) 3.60 (3.50-3.71) <.001 2.97 (2.88-3.07) <.001 3.16 (3.06-3.26) <.001 2.76 (2.62-2.89) <.001
Sandwich restaurant
6-in Tuna 4.49 (4.40-4.59) 4.62 (4.54-4.70) .04 4.80 (4.72-4.88) <.001 4.63 (4.55-4.72) .03 4.73 (4.65-4.82) <.001
Grilled chicken salad 4.63 (4.54-4.72) 4.81 (4.73-4.89) .004 4.93 (4.87-5.00) <.001 4.81 (4.73-4.89) .005 4.88 (4.79-4.96) <.001
Roast beef salad 4.39 (4.29-4.48) 4.28 (4.19-4.37) .11 3.45 (3.36-3.54) <.001 3.58 (3.49-3.67) <.001 3.74 (3.63-3.84) <.001
12-in Steak and cheese 4.17 (4.06-4.27) 3.93 (3.84-4.02) .001 3.22 (3.13-3.31) <.001 3.39 (3.29-3.48) <.001 2.98 (2.85-3.12) <.001
Open in a new tab
a

Mean predicted scores from postestimation margins obtained from weighted simple linear regression.

b

Rated on a 7-point scale: 1 = very unappealing; 2 = unappealing; 3 = slightly unappealing; 4 = neither appealing nor unappealing; 5 = slightly appealing; 6 = appealing; 7 = very appealing.

c

Rated on a 7-point scale: 1 = very unhealthy; 2 = unhealthy; 3 = slightly unhealthy; 4 = neither healthy nor unhealthy; 5 = slightly healthy; 6 = healthy; 7 = very healthy.

d

Rated on a 7-point scale: 1 = extremely high climate impact; 2 = very high; 3 = high; 4 = moderate; 5 = low; 6 = very low; 7 = extremely low.

Compared to the control label, all ecolabels were less believable but had higher PME, and only the high-impact and the TLL were ranked higher for grabbing attention (Table 3). The TLL was ranked highest for believability (3.11 [95% CI, 3.03-3.19]) other than the control label (3.56 [95% CI, 3.48-3.63]) and had the highest PME of all the labels (2.62 [95% CI, 2.54-2.71]).

Table 3. Mean Predicted Attitudes and Beliefs About Labels by Treatment Condition (N = 6210)a,b.

Attitude/belief Control, mean (95% CI) (n = 1236) Low climate impact (n = 1249) Traffic light (n = 1235) High climate impact (n = 1243) Grade scale (n = 1247)
Mean (95% CI) P value Mean (95% CI) P value Mean (95% CI) P value Mean (95% CI) P value
This label is believable to me. 3.56 (3.48-3.63) 2.80 (2.71-2.88) <.001 3.11 (3.03-3.19) <.001 2.99 (2.90-3.08) <.001 3.05 (2.96-3.14) <.001
This label grabs my attention. 2.96 (2.88-3.04) 2.74 (2.66-2.83) <.001 3.20 (3.12-3.29) <.001 3.21 (3.12-3.30) <.001 2.99 (2.90-3.08) .678
Mean perceived message effectivenessc 1.88 (1.80-1.95) 2.22 (2.15-2.29) <.001 2.62 (2.54-2.71) <.001 2.60 (2.51-2.69) <.001 2.56 (2.48-2.65) <.001
This label discourages me from wanting to eat foods with a high climate impact. 1.82 (1.74-1.91) 2.11 (2.03-2.20) <.001 2.53 (2.43-2.62) <.001 2.54 (2.44-2.63) <.001 2.48 (2.39-2.58) <.001
This label makes eating foods with a high climate impact seem unpleasant. 1.84 (1.76-1.92) 2.15 (2.06-2.23) <.001 2.61 (2.51-2.70) <.001 2.63 (2.53-2.73) <.001 2.57 (2.47-2.67) <.001
This label makes me concerned about the health effects of eating menu items with a high climate impact. 1.95 (1.87-2.04) 2.25 (2.16-2.34) <.001 2.62 (2.52-2.71) <.001 2.55 (2.45-2.65) <.001 2.56 (2.46-2.65) <.001
This label makes me concerned about the environmental effects of eating menu items with a high climate impact. 1.89 (1.80-1.98) 2.37 (2.28-2.46) <.001 2.74 (2.65-2.83) <.001 2.68 (2.58-2.78) <.001 2.65 (2.55-2.75) <.001
Open in a new tab
a

Mean predicted scores from postestimation margins obtained from weighted simple linear regression.

b

Rated on a 5-point scale: 1 = not at all; 2 = a little bit; 3 = somewhat; 4 = quite a bit; 5 = a great deal.

c

Mean score from the 4 Perceived Message Effectiveness questions.

Participants in all ecolabel conditions were more likely to correctly identify menu items with higher climate impact compared to the control with the highest proportion of correct responses in the TLL condition (66.07% correct [95% CI, 62.64%-69.50%]) (eTable 6 in Supplement 2). There were no differences in probability of selecting a sugar-sweetened beverage (eFigure 5 in Supplement 2), and no differences in the overall cost of the meals selected (eFigure 6 in Supplement 2) by label condition in either restaurant.

Discussion

In this web-based randomized clinical trial with a large, nationally representative sample, we found that TLL-style ecolabels in which items receive a red, yellow, or green label led to more healthful meal selections in a burger-focused fast-food restaurant and lower climate impact (GHGE) meal selections in both burger- and sandwich-focused restaurants. While the magnitude of differences in the healthfulness of meal selections compared to the control and other ecolabels was small, and less than 50% of participants said they used the labels, the TLL consistently and meaningfully performed better than the control and other ecolabels across multiple outcomes related to sustainability, label noticeability and understanding, knowledge about healthfulness and climate impact. Meal cost did not differ across label conditions. Results signal that TLL-style menu ecolabels may be effective at promoting more sustainable food choices in fast-food restaurants while also educating consumers about how food choices contribute to climate change. TLLs may also promote more healthful food choices in some restaurants without incurring extra costs to consumers or adversely impacting fast-food restaurants’ profits.

These results align with prior research showing the effectiveness of interpretive food/nutrition labels,30,32 TLL (that provide information about high- and low-impact items simultaneously) being most effective, and showing small effects of ecolabels on nutrition outcomes.25,26,30 Compared to the control condition, the TLL led to 15.2% and 29.8% lower GHGE for meals ordered from the burger restaurant and sandwich restaurant, respectively. Given the scale and popularity of large chain restaurants,13,14 reductions of GHGE of this magnitude could make a meaningful cumulative difference in addressing climate change and should be considered as part of multifaceted strategies to promote more sustainable food industry practices and consumer choices.

While consumers have indicated that sustainable food choices are important to them, public awareness of the role of reducing red meat intake to mitigate climate change is low.48,49 The 2025 to 2030 Dietary Guidelines for Americans’ increased emphasis on promoting red meat may shape public perceptions and complicate efforts to align dietary behaviors with sustainability goals.50 The present results signal that ecolabels on restaurant menus may increase knowledge and awareness about the climate impact of menu items, particularly red meat items compared to other protein sources, even though fewer than half used the labels in their food selections, despite noticing and accurately reporting what the labels conveyed. Ultimately, ecolabels may impact consumer behavior and food choices at the time of purchase but also in the future and, perhaps, in other venues beyond fast-food restaurants19,24,25,33; however, more research is needed in retail and restaurant settings and with repeated exposures to labels to explore these possibilities further.

Whether ecolabel effects would vary depending on restaurant type has been a key question. We found that, in both restaurants, the TLL led to more sustainable food selections. However, we found differential effects on the healthfulness of selections. As in prior research, differences in meal healthfulness scores and nutrition outcomes were small in both restaurants, likely due, in part, to the limited range of nutrition values across menu items.25,26,30 It will be important for future research to investigate whether ecolabels could have a stronger effect on meal selection healthfulness when tested in restaurants with greater variation in menu items’ nutritional profiles.

Food and menu labels are a feasible policy that can prompt supply-side changes via reformulation or menu composition and, in some instances, can shift consumer knowledge and behavior.31,33,51,52,53,54,55,56 Ecolabels may use these same mechanisms to reduce food systems-related climate change. Internationally, ecolabels have garnered interest from governments.57,58 In the US, though there had previously been some federal interest in pursuing climate food labels,59 there is currently more promise for state- or local-level policy. Further, in the absence of local, state, or federal policy, restaurants and food companies can take voluntary action to implement ecolabels, which some have already done.60,61,62 Mandatory labels, with a consistent, evidence-based design, would be helpful to limit consumer confusion and prompt greater industry and consumer behavior change.

Limitations

The primary limitation of this study is that it was a hypothetical, 1-time ordering task simulation in which participants did not order food for actual consumption and did not spend their own money, which limits generalizability. Although the study’s purpose was masked, social desirability bias may still be a concern. As with other nutrition scoring systems, the NPI measure has limitations. NPI scores are applied at the item, rather than ingredient or meal level. Additionally, NPI scoring does not differentiate beverages well. The modified meal-level NPI (weighted mean of food item scores), attempts to address these limitations. The limited range of menu items and nutrition values across menu items, and the per–100-g basis of the primary outcome (NPI score), which does not differentiate portion sizes, may have limited our ability to see differences in these outcomes. Additionally, while our calculations of climate impact focus on GHGE, impacts on biodiversity, water use, soil health, and pollution are also important sustainability considerations. Future research is needed to test ecolabels in restaurants with greater heterogeneity in nutrition and climate impact of menu items.

Conclusions

This randomized clinical trial demonstrated that traffic-light–style menu ecolabels may be an effective strategy to promote more sustainable food choices without increasing consumer costs. In some restaurants, ecolabels may also lead to more healthful food choices; however, more research is needed, particularly in different types of restaurants.

Supplement 1.

Trial Protocol

jamahealthforum-e262108-s001.pdf (348.3KB, pdf)
Supplement 2.

eAppendix 1. Nutrition and Price Information Methods Appendix

eAppendix 2. Greenhouse Gas Emissions Methods Appendix

eAppendix 3. Results Appendix.

eTable 1. Nutrition Profile Index, Calories, and Greenhouse Gas Emissions for Menu Items Shown to Participants

eTable 2. Secondary Outcomes wording and response options

eTable 3. Menu Item Selection Frequency (n [%]) by Experimental Condition (N=6210)

eTable 4. Mean Predicted Nutrients of Full Meal Ordered by Experimental Condition and Restaurant (N=6210)

eFigure 1. Mean Predicted Nutrition Profile Index Score of Meal Selected at the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Experimental Condition and Noticeability (n=6209)

eTable 5. Mean Predicted Probabilities of Noticing Label, Identifying What Label was About, and Label Use by Treatment Condition

eTable 6. Mean Predicted Probabilities of Correctly Identifying Higher Climate Impact Menu Items and Ranking Items by Climate Impact by Treatment Condition (n=6242)

eFigure 2. Mean Predicted Nutrition Profile Index Score of Selected Main Item at the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Treatment Condition (n=6210)

eFigure 3. Predicted Probability of Selecting Sustainable Item at the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Treatment Condition (n=6210)

eFigure 4. Predicted Probability of Selecting Item Containing Red Meat the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Treatment Condition (n=6210)

eFigure 5. Predicted Probability of Selecting a Sugar Sweetened Beverage at the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Treatment Condition (n=6210)

eFigure 6. Mean Predicted Cost of selected Meals at the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Treatment Condition (n=6210)

eTable 7. Summary Table of Key Results Across Label Conditions

eAppendix 4. Menu images for different label conditions

jamahealthforum-e262108-s002.pdf (11.3MB, pdf)
Supplement 3.

Data Sharing Statement

jamahealthforum-e262108-s003.pdf (17KB, pdf)

References

  • 1.Crippa M, Solazzo E, Guizzardi D, Monforti-Ferrario F, Tubiello FN, Leip A. Food systems are responsible for a third of global anthropogenic GHG emissions. Nat Food. 2021;2(3):198-209. doi: 10.1038/s43016-021-00225-9 [DOI] [PubMed] [Google Scholar]
  • 2.Rockström J, Thilsted SH, Willett WC, et al. The EAT-Lancet Commission on healthy, sustainable, and just food systems. Lancet. 2025;406(10512):1625-1700. doi: 10.1016/S0140-6736(25)01201-2 [DOI] [PubMed] [Google Scholar]
  • 3.Poore J, Nemecek T. Reducing food’s environmental impacts through producers and consumers. Science. 2018;360(6392):987-992. doi: 10.1126/science.aaq0216 [DOI] [PubMed] [Google Scholar]
  • 4.Pan A, Sun Q, Bernstein AM, et al. Red meat consumption and risk of type 2 diabetes: 3 cohorts of US adults and an updated meta-analysis. Am J Clin Nutr. 2011;94(4):1088-1096. doi: 10.3945/ajcn.111.018978 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Larsson SC, Orsini N. Red meat and processed meat consumption and all-cause mortality: a meta-analysis. Am J Epidemiol. 2014;179(3):282-289. doi: 10.1093/aje/kwt261 [DOI] [PubMed] [Google Scholar]
  • 6.Yang C, Pan L, Sun C, Xi Y, Wang L, Li D. Red meat consumption and the risk of stroke: a dose-response meta-analysis of prospective cohort studies. J Stroke Cerebrovasc Dis. 2016;25(5):1177-1186. doi: 10.1016/j.jstrokecerebrovasdis.2016.01.040 [DOI] [PubMed] [Google Scholar]
  • 7.Zheng Y, Li Y, Satija A, et al. Association of changes in red meat consumption with total and cause specific mortality among US women and men: two prospective cohort studies. BMJ. 2019:l2110. doi: 10.1136/bmj.l2110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Dong X, Zhuang Z, Zhao Y, et al. Unprocessed red meat and processed meat consumption, plasma metabolome, and risk of ischemic heart disease: a prospective cohort study of UK Biobank. J Am Heart Assoc. 2023;12(7):e027934. doi: 10.1161/JAHA.122.027934 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Davies KP, Gibney ER, Leonard UM, et al. Sustainable diets reduce diet-related greenhouse gas emissions and improve diet quality: results from the MyPlanetDiet randomized controlled trial. Am J Clin Nutr. 2025;122(4):962-971. doi: 10.1016/j.ajcnut.2025.05.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Willits-Smith A, Aranda R, Heller MC, Rose D. Addressing the carbon footprint, healthfulness, and costs of self-selected diets in the USA: a population-based cross-sectional study. Lancet Planet Health. 2020;4(3):e98-e106. doi: 10.1016/S2542-5196(20)30055-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Rose D, Heller MC, Willits-Smith AM, Meyer RJ. Carbon footprint of self-selected US diets: nutritional, demographic, and behavioral correlates. Am J Clin Nutr. 2019;109(3):526-534. doi: 10.1093/ajcn/nqy327 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Nelson ME, Hamm MW, Hu FB, Abrams SA, Griffin TS. Alignment of healthy dietary patterns and environmental sustainability: a systematic review. Adv Nutr. 2016;7(6):1005-1025. doi: 10.3945/an.116.012567 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Fryar CD, Hughes JP, Herrick KA, Ahluwalia N. Fast food consumption among adults in the United States, 2013–2016. NCHS Data Brief. 2018;(322):1-8. https://www.cdc.gov/nchs/data/databriefs/db322-h.pdf [PubMed]
  • 14.Dunn CG, Gao KJ, Soto MJ, Bleich SN. Disparities in adult fast-food consumption in the U.S. by race and ethnicity, National Health and Nutrition Examination Survey 2017-2018. Am J Prev Med. 2021;61(4):e197-e201. doi: 10.1016/j.amepre.2021.01.043 [DOI] [PubMed] [Google Scholar]
  • 15.Frank SM, Taillie LS, Jaacks LM. How Americans eat red and processed meat: an analysis of the contribution of thirteen different food groups. Public Health Nutr. 2022;25(5):1406-1415. doi: 10.1017/S1368980022000416 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.National Restaurant Association . What’s hot in 2022?—health, happiness, plus a side of sustainability. National Restaurant Association. Accessed December 14, 2025. https://restaurant.org/education-and-resources/resource-library/whats-hot-in-2022/
  • 17.National Restaurant Association . Sustainability, health and variety top list of 2025 macro trends: operators are adding more environmental and personal well-being options to attract a greater diversity of diners. Accessed December 14, 2025. https://restaurant.org/education-and-resources/resource-library/sustainability,-health-and-variety-top-list-of-2025-macro-trends/
  • 18.DiPalma B. From Burger King to Panda Express, here’s who’s riding the plant-based bandwagon. Yahoo!finance. Accessed January 5, 2026. https://finance.yahoo.com/news/from-burger-king-to-panda-express-heres-whos-on-the-plant-based-bandwagon-183203937.html
  • 19.Asioli D, Aschemann-Witzel J, Nayga RM. Sustainability-related food labels. Annu Rev Resour Economics. 2020;12(1):171-185. doi: 10.1146/annurev-resource-100518-094103 [DOI] [Google Scholar]
  • 20.Kateman B. Carbon labels are finally coming to the food and beverage industry. Forbes. Accessed January 5, 2026. https://www.forbes.com/sites/briankateman/2020/07/20/carbon-labels-are-finally-coming-to-the-food-and-beverage-industry/?sh=59a2731c7c03
  • 21.Liu J, Rehm CD, Micha R, Mozaffarian D. Quality of meals consumed by US adults at full-service and fast-food restaurants, 2003-2016: persistent low quality and widening disparities. J Nutr. 2020;150(4):873-883. doi: 10.1093/jn/nxz299 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Vinyard M, Zimmer M, Herrick KA, Story M, Juan W, Reedy J. Healthy Eating Index-2015 scores vary by types of food outlets in the United States. Nutrients. 2021;13(8):2717. doi: 10.3390/nu13082717 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Nagao-Sato S, Reicks M. Food away from home frequency, diet quality, and health: cross-sectional analysis of NHANES Data 2011-2018. Nutrients. 2022;14(16):3386. doi: 10.3390/nu14163386 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Potter C, Bastounis A, Hartmann-Boyce J, et al. The effects of environmental sustainability labels on selection, purchase, and consumption of food and drink products: a systematic review. Environ Behav. 2021;53(8):891-925. doi: 10.1177/0013916521995473 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Grummon AH, Zeitlin AB, Lee CJY, et al. Ecolabels and the healthfulness and carbon footprint of restaurant meal selections: a randomized clinical trial. JAMA Netw Open. 2025;8(8):e2524773. doi: 10.1001/jamanetworkopen.2025.24773 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Wolfson JA, Musicus AA, Leung CW, Gearhardt AN, Falbe J. Effect of climate change impact menu labels on fast food ordering choices among US adults: a randomized clinical trial. JAMA Netw Open. 2022;5(12):e2248320-e2248320. doi: 10.1001/jamanetworkopen.2022.48320 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Taillie LS, Wolfson JA, Prestemon CE, et al. The impact of an eco-score label on US consumers’ perceptions of environmental sustainability and intentions to purchase food: a randomized experiment. PLoS One. 2024;19(6):e0306123. doi: 10.1371/journal.pone.0306123 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Potter C, Pechey R, Clark M, et al. Effects of environmental impact labels on the sustainability of food purchases: a randomised controlled trial in an experimental online supermarket. PLoS One. 2024;19(9):e0309386. doi: 10.1371/journal.pone.0309386 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Arrazat L, Chambaron S, Arvisenet G, et al. Traffic-light front-of-pack environmental labelling across food categories triggers more environmentally friendly food choices: a randomised controlled trial in virtual reality supermarket. Int J Behav Nutr Phys Act. 2023;20(1):7. doi: 10.1186/s12966-023-01410-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Reimold AE, Falbe J, Kim BF, et al. The impact of restaurant menu eco labels on consumer meal selections: an RCT. Am J Prev Med. 2026;70(6):108265. doi: 10.1016/j.amepre.2026.108265 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Cecchini M, Warin L. Impact of food labelling systems on food choices and eating behaviours: a systematic review and meta-analysis of randomized studies. Obes Rev. 2016;17(3):201-210. doi: 10.1111/obr.12364 [DOI] [PubMed] [Google Scholar]
  • 32.Grummon AH, Gibson LA, Musicus AA, Stephens-Shields AJ, Hua SV, Roberto CA. Effects of 4 interpretive front-of-package labeling systems on hypothetical beverage and snack selections: a randomized clinical trial. JAMA Netw Open. 2023;6(9):e2333515. doi: 10.1001/jamanetworkopen.2023.33515 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Roberto CA, Ng SW, Ganderats-Fuentes M, et al. The influence of front-of-package nutrition labeling on consumer behavior and product reformulation. Annu Rev Nutr. 2021;41(1):529-550. doi: 10.1146/annurev-nutr-111120-094932 [DOI] [PubMed] [Google Scholar]
  • 34.Besson T, Bouxom H, Jaubert T. Halo It’s meat! the effect of the vegetarian label on calorie perception and food choices. Ecol Food Nutr. 2020;59(1):3-20. doi: 10.1080/03670244.2019.1652820 [DOI] [PubMed] [Google Scholar]
  • 35.Chandon P, Wansink B. The biasing health halos of fast-food restaurant health claims: lower calorie estimates and higher side-dish consumption intentions. J Consum Res. 2007;34(3):301-314. doi: 10.1086/519499 [DOI] [Google Scholar]
  • 36.NORC at the University of Chicago . Technical overview of the AmeriSpeak panel NORC’s probability-based household panel. Accessed June 11, 2026. https://www.norc.org/services-solutions/amerispeak.html
  • 37.Wolfson JA, Leung CW, Gearhardt AN. Trends in the nutrition profile of menu items at large burger chain restaurants. Am J Prev Med. 2020;58(6):e171-e179. doi: 10.1016/j.amepre.2020.01.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Lesser LI, Wu L, Matthiessen TB, Luft HS. Evaluating the healthiness of chain-restaurant menu items using crowdsourcing: a new method. Public Health Nutr. 2017;20(1):18-24. doi: 10.1017/S1368980016001804 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Harris JL, Schwartz MB, Munsell CR, et al. Fast Food FACTS 2013: Measuring Progress in Nutrition and Marketing to Children and Teens. Yale Rudd Center for Food Policy and Obesity; 2013. [Google Scholar]
  • 40.Scarborough P, Boxer A, Rayner M, Stockley L. Testing nutrient profile models using data from a survey of nutrition professionals. Public Health Nutr. 2007;10(4):337-345. doi: 10.1017/S1368980007666671 [DOI] [PubMed] [Google Scholar]
  • 41.Arambepola C, Scarborough P, Rayner M. Validating a nutrient profile model. Public Health Nutr. 2008;11(4):371-378. doi: 10.1017/S1368980007000377 [DOI] [PubMed] [Google Scholar]
  • 42.Rayner M, Scarborough P, Boxer A, Stockley L. Nutrient Profiles: Development of Final Model. Food Standards Agency; 2005. [Google Scholar]
  • 43.Waite R, Vennard D, Pozzi G. Tracking Progress Towards the Cool Food Pledge: Setting Climate Targets, Tracking Metrics, Using the Cool Food Calculator, and Related Guidance for Pledge Signatories; Technical Note. World Resources Institute; 2020. [Google Scholar]
  • 44.World Resources Institute . Cool food. Accessed June 30, 2021. https://www.wri.org/initiatives/cool-food-pledge
  • 45.Waite R, Blondin S. Identifying cool food meals. World Resources Institute. Updated June 27, 2022. Accessed April 8, 2026. https://www.wri.org/research/identifying-cool-food-meals
  • 46.Baig SA, Noar SM, Gottfredson NC, Boynton MH, Ribisl KM, Brewer NT. UNC Perceived Message Effectiveness: validation of a brief scale. Ann Behav Med. 2019;53(8):732-742. doi: 10.1093/abm/kay080 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Noar SM, Barker J, Bell T, Yzer M. Does Perceived Message Effectiveness predict the actual effectiveness of tobacco education messages?—a systematic review and meta-analysis. Health Commun. 2020;35(2):148-157. doi: 10.1080/10410236.2018.1547675 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Slotnick MJ, Falbe J, Cohen JFW, Gearhardt AN, Wolfson JA, Leung CW. Environmental and climate impact perceptions in university students: sustainability motivations and perceptions correspond with lower red meat intake. J Acad Nutr Diet. 2023;123(5):740-750. doi: 10.1016/j.jand.2022.09.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Slotnick MJ, Falbe J, Wolfson JA, Jones AD, Leung CW. Environmental-, climate-, and health-related dietary motivations are associated with higher diet quality in a national sample of US adults with lower incomes. J Acad Nutr Diet. 2024;124(5):594-606. doi: 10.1016/j.jand.2023.11.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.US Department of Agriculture . Dietary guidelines for Americans, 2025-2030. 2026. http://realfood.gov/
  • 51.Bleich SN, Wolfson JA, Jarlenski MP, Block JP. Restaurants with calories displayed on menus had lower calorie counts compared to restaurants without such labels. Health Aff (Millwood). 2015;34(11):1877-1884. doi: 10.1377/hlthaff.2015.0512 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Petimar J, Ramirez M, Rifas-Shiman SL, et al. Evaluation of the impact of calorie labeling on McDonald’s restaurant menus: a natural experiment. Int J Behav Nutr Phys Act. 2019;16(1):99. doi: 10.1186/s12966-019-0865-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Grummon AH, Petimar J, Soto MJ, et al. Changes in calorie content of menu items at large chain restaurants after implementation of calorie labels. JAMA Netw Open. 2021;4(12):e2141353. doi: 10.1001/jamanetworkopen.2021.41353 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Petimar J, Zhang F, Rimm EB, et al. Changes in the calorie and nutrient content of purchased fast food meals after calorie menu labeling: a natural experiment. PLoS Med. 2021;18(7):e1003714. doi: 10.1371/journal.pmed.1003714 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Rummo PE, Mijanovich T, Wu E, et al. Menu Labeling and calories purchased in restaurants in a US national fast food chain. JAMA Netw Open. 2023;6(12):e2346851. doi: 10.1001/jamanetworkopen.2023.46851 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Shrestha A, Cullerton K, White KM, Mays J, Sendall M. Impact of front-of-pack nutrition labelling in consumer understanding and use across socio-economic status: A systematic review. Appetite. 2023;187:106587. doi: 10.1016/j.appet.2023.106587 [DOI] [PubMed] [Google Scholar]
  • 57.Southey F. Denmark ‘first country in the world’ to develop its own climate label for food. Food Navigator Europe. April 19, 2022. Accessed January 5, 2026. https://www.foodnavigator.com/Article/2022/04/19/denmark-first-country-in-the-world-to-develop-its-own-climate-label-for-food/
  • 58.Coyne A. Food giants sign up to new Foundation Earth eco-labelling scheme. JustFood. June 27, 2021. Accessed January 5, 2026. https://www.just-food.com/news/food-giants-sign-up-to-new-foundation-earth-eco-labelling-scheme/
  • 59.Brownley J. Brownley, Casten introduce legislation to inform consumers with carbon footprint labels for food products. Accessed January 5, 2026. https://juliabrownley.house.gov/brownley-casten-introduce-legislation-to-inform-consumers-with-carbon-footprint-labels-for-food-products/
  • 60.Aramark. Aramark and Drexel Food Lab collaborate to advance sustainable food practices. Accessed February 5, 2026. https://www.aramark.com/newsroom/news/2024/november/drexel-food-lab
  • 61.Hospitality Technology. Panera Bread labels ‘climate-friendly’ menu items. Accessed February 5, 2026. https://hospitalitytech.com/panera-bread-labels-climate-friendly-menu-items
  • 62.Just Salad . Big flavor, low impact—why we carbon label. Accessed Februrary 5, 2026. https://www.justsalad.com/carbonlabel/

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1.

Trial Protocol

jamahealthforum-e262108-s001.pdf (348.3KB, pdf)
Supplement 2.

eAppendix 1. Nutrition and Price Information Methods Appendix

eAppendix 2. Greenhouse Gas Emissions Methods Appendix

eAppendix 3. Results Appendix.

eTable 1. Nutrition Profile Index, Calories, and Greenhouse Gas Emissions for Menu Items Shown to Participants

eTable 2. Secondary Outcomes wording and response options

eTable 3. Menu Item Selection Frequency (n [%]) by Experimental Condition (N=6210)

eTable 4. Mean Predicted Nutrients of Full Meal Ordered by Experimental Condition and Restaurant (N=6210)

eFigure 1. Mean Predicted Nutrition Profile Index Score of Meal Selected at the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Experimental Condition and Noticeability (n=6209)

eTable 5. Mean Predicted Probabilities of Noticing Label, Identifying What Label was About, and Label Use by Treatment Condition

eTable 6. Mean Predicted Probabilities of Correctly Identifying Higher Climate Impact Menu Items and Ranking Items by Climate Impact by Treatment Condition (n=6242)

eFigure 2. Mean Predicted Nutrition Profile Index Score of Selected Main Item at the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Treatment Condition (n=6210)

eFigure 3. Predicted Probability of Selecting Sustainable Item at the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Treatment Condition (n=6210)

eFigure 4. Predicted Probability of Selecting Item Containing Red Meat the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Treatment Condition (n=6210)

eFigure 5. Predicted Probability of Selecting a Sugar Sweetened Beverage at the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Treatment Condition (n=6210)

eFigure 6. Mean Predicted Cost of selected Meals at the Burger Restaurant (Panel A) and the Sandwich Restaurant (Panel B) by Treatment Condition (n=6210)

eTable 7. Summary Table of Key Results Across Label Conditions

eAppendix 4. Menu images for different label conditions

jamahealthforum-e262108-s002.pdf (11.3MB, pdf)
Supplement 3.

Data Sharing Statement

jamahealthforum-e262108-s003.pdf (17KB, pdf)

Articles from JAMA Health Forum are provided here courtesy of American Medical Association

RESOURCES