Skip to main content
NCBI home page
JAMA Network logoLink to JAMA Network
. 2026 Jun 23;9(6):e2619650. doi: 10.1001/jamanetworkopen.2026.19650

Environmental Menu Labels and Food Selection Among US Adults

A Randomized Clinical Trial

Alexandria E Reimold 1, Julia A Wolfson 2,3, Brittany Lemmon 4, Aviva A Musicus 5,6, Brent F Kim 7,8, Nina M Carr 2, Raychel Santo 9, Clara Cho 9, Cindy W Leung 5, Jennifer Falbe 1,✉
PMCID: PMC13291850  PMID: 42334851

Key Points

Question

Do climate labels on fast food menus affect the sustainability and nutritional quality of food selections?

Findings

In an online randomized clinical trial of 10 342 US participants, individuals were significantly less likely to select foods with high climate impact or greenhouse gas emissions when menu items had climate grade or high climate impact labels compared with the control condition. Menu items with a climate grade label displaying a scale of all possible grades resulted in, on average, participants selecting foods with lower saturated fat compared with the control condition.

Meaning

These findings suggest that displaying climate labels on menu items, particularly labels with climate grade plus scale designs, may reduce climate impact and improve nutrition of fast food selections.

This randomized clinical trial examines the potential for climate labels on fast food menus to encourage more sustainable and healthier food choices.

Abstract

Importance

Food systems account for approximately one-third of global greenhouse gas emissions (GHGE), and many foods with high GHGE pose risks to human health. Environmental impact (eg, climate) labels on restaurant menus may promote lower GHGE and higher nutritional quality food choices.

Objective

To compare the efficacy of climate labels on the GHGE and nutritional quality of food choices.

Design, Setting, and Participants

This online randomized clinical trial was conducted using a menu based on a US fast food chain between December 2024 and February 2025. Participants included US adults aged 18 years or older. Analyses were conducted between February and April 2025.

Intervention

Participants were randomized to 1 of 6 label conditions: (1) control, (2) environmental cost (GHGE represented as a monetary cost), (3) climate grade (GHGE represented as a grade [eg, A to F, with A representing the lowest GHGE]), (4) climate grade plus scale (GHGE represented as a grade accompanied by a scale showing all possible grades), (5) high climate impact on items with high GHGE, and (6) warning high climate impact on items with high GHGE. Participants ordered from a fast food menu displaying their assigned label next to applicable items.

Main Outcomes and Measures

Primary outcomes include the selection of an item with high climate impact and GHGE of menu orders. Secondary outcomes include Nutrient Profile Index (NPI) score and nutrients of menu orders and participant perceptions.

Results

This study included 10 342 participants (mean [SD] age: 46.4 [16.9] years; 53% women). Participants in the climate grade (825 of 1721 [48%]; P = .02), climate grade plus scale (787 of 1719 [46%]; P < .001), high climate impact (777 of 1724 [45%]; P < .001), and warning high climate impact (817 of 1725 [47%]; P = .01) conditions were significantly less likely to select an item with a high climate impact than those in the control condition (920 of 1728 [53%]). Participants in the climate grade plus scale (mean GHGE, 16.3 [95% CI, 15.5-17.2] kg CO2 equivalent, adjusted P < .001) and high climate impact (mean GHGE, 16.9 [95% CI, 16.0-17.8] kg CO2 equivalent, adjusted P = .04) conditions selected items with lower GHGE than those in the control condition (mean GHGE, 18.8 [95% CI, 17.9-19.6] kg CO2 equivalent). NPI scores did not differ among label conditions, but participants in the climate grade plus scale condition selected items with less saturated fat than those in the control condition (11.8 g vs 12.6 g; adjusted P = .04). Although there were no significant differences between the environmental cost and control labels for primary or secondary nutrition outcomes, when viewing labels in isolation, participants perceived the environmental cost label as the most discouraging of selecting high climate impact items (mean score, 3.0 [95% CI, 2.9-3.1]; P < .001).

Conclusions and Relevance

This randomized clinical trial’s findings suggest that using climate labels on restaurant menus, particularly climate grade plus scale designs, may decrease climate impact and improve nutrition of restaurant orders. Future research is needed on environmental cost label designs and climate label use in real restaurant settings.

Trial Registration

ClinicalTrials.gov Identifier: NCT06651060

Introduction

Individuals have expressed interest in reducing their climate impact but are unsure how.1,2,3,4,5 Because food systems account for roughly one-third of anthropogenic global greenhouse gas emissions (GHGE),6 food purchasing behaviors provide an opportunity for intervention. In particular, red meat produces significantly more GHGE than other proteins7 and is associated with increased risk of type 2 diabetes, cardiovascular disease, cancer, and early mortality.8,9 Fast food accounts for roughly 20% of red meat intake in the US,10 so reducing GHGE from fast food orders may improve both planetary and human health.

Labels indicating the environmental impact of menu items (climate labels) can significantly decrease GHGE of meals ordered and improve meal nutrition.11,12 Labels that grade items’ climate impact (eg, European Green-score label [formerly Eco-score])13 or warn about items with high climate impact (similar to sodium warning labels)11,12,14 may be particularly effective at encouraging selections of menu items with low GHGE. However, ideal label designs have yet to be identified. For example, climate grade labels with and without a scale of possible grades and warning labels with and without a marker word (Warning) have not been compared on fast food menus.

Additionally, True Cost Accounting (TCA) is an innovative method used for calculating comprehensive prices of commodities by accounting for associated environmental and social costs.15,16,17 The state of New York is piloting TCA methods for public agency food procurement,18,19 and global organizations are working to understand the true cost of food.20 Yet, no study to date has assessed TCA as part of a climate label design. To determine the efficacy of climate grade, warning, and environmental cost labels on the climate impact and nutritional quality of fast food menu orders, we conducted an online randomized clinical trial with a large sample of US adults who regularly consume fast food.

Methods

Participants

From December 2024 to February 2025, we used an online survey recruitment platform, CloudResearch, to recruit participants from the research panel Prime Panels. Eligible participants were US residents aged 18 years or older who purchased food from any restaurant at least once per month. We used quota sampling based on the 2023 US Census American Community Survey 5-year estimates to ensure that sample demographics (gender, race, ethnicity, age, educational attainment) were representative of the US population. Participants provided electronic informed consent by reviewing study information and then selecting yes when asked if they would like to continue (trial protocol in Supplement 1). The University of California, Davis institutional review board approved this study. We followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guidelines. Before data collection, we preregistered the target sample size and analysis plan for primary and secondary outcomes.

Procedures

Before randomization, all participants responded to questions for an unrelated parent study on added-sugar labels that used different menus from the current study. Participants were then randomly assigned to 1 of 6 label conditions: (1) control (a QR code linked to additional menu information), (2) environmental cost on all items, (3) climate grade on all items, (4) climate grade plus scale on all items, (5) high climate impact on items with high GHGE, and (6) warning high climate impact on items with high GHGE (eFigure in Supplement 2). Labels contained both text and an icon (control labels had a QR code and climate labels had a globe) and were based on prior studies.11,12

Participants viewed their assigned label below items on a realistic, 16-item menu modeled after an online menu from a popular US fast food chain, including a variety of beef burgers, chicken and fish items, salads, and a plant-based burger. Animal proteins are main drivers of food-related GHGE, and entrées generally had 1 to 2 orders of magnitude greater emissions than sides, beverages, and desserts, so we limited the menu to entrées only. Each menu displayed a label description, prices, and calories for each item. Menu item price, calorie, ingredient, and nutrition information was extracted from the chain’s website. We calculated mean prices (across 13 US cities in October 2024), then rounded up to the nearest value ending in $0.09 to match the chain’s pricing scheme. An example menu is shown in eFigure in Supplement 2.

We summed the supply chain GHGE and carbon opportunity cost in kilograms of carbon dioxide equivalent (kg CO2e). Carbon opportunity cost assumes global warming potential over 100 years per kilogram of each menu item ingredient using data from the World Resources Institute (WRI) Coolfood Pledge Coolfood Meals calculator.21 We calculated environmental cost labels with the True Price monetization factor that converts GHGE to US dollars.17 Label thresholds were calculated using methods from Waite and Blondin,22 assessing items relative to the average regional meal-related GHGE and reductions needed to align with climate mitigation goals (eAppendix in Supplement 2).

Similar to previously published menu label studies, participants were instructed to order 1 item to reduce participant burden.11 To encourage realistic menu selections, participants were informed that 20 randomly selected participants could opt to receive their selected menu item. Participants then responded to questions about their assigned label, related behaviors, and personal opinions. All participants received compensation from CloudResearch.

Measures

Primary outcomes were (1) a dichotomous indicator of whether participants selected an item with a high climate impact, and (2) a continuous measure of total GHGE for the item selected. Secondary outcomes included a continuous Nutrient Profile Index (NPI) score, which indicates nutritional quality, and continuous measures of calories (in kilocalories [kcal]), saturated fat (in grams [g]), and sodium (in milligrams [mg]) of the selected item. NPI scores combine positive points for fruits, vegetables, nuts, and fiber, and negative points for calories, saturated fat, sodium, and sugar.23

Additional outcomes included whether participants noticed the label, the degree to which they would support the label as a policy option, and the degree to which participants perceived the label to discourage them from purchasing an item with a high climate impact (ie, perceived discouragement). Similar to previous research, perceived discouragement was measured with 1 item from the University of North Carolina Perceived Message Effectiveness Scale.11,24,25 Self-reported participant behaviors, opinions, and demographic characteristics were evaluated with additional survey questions. Exact item wording is provided in eTable 1 in Supplement 2.

Statistical Analysis

Analyses were conducted between February and April 2025. The sample size was powered for the unrelated parent study; however, post hoc power analysis using G*Power, version 3.1.9.5 (University of Dusseldorf) indicated that this study had statistical power greater than 99% (assuming α = .05) to detect a small effect (Cohen d = 0.20) when comparing the intervention conditions to the control condition and to each other. We summarize participant characteristics and menu item category order frequency with descriptive statistics. We used ordinary least-squares linear regression models to evaluate the impact of labels on continuous outcomes and Poisson regression with robust error variance26 for dichotomous outcomes, regressing outcomes on indicators for experimental condition. We first compared each climate label to the control label and then conducted pairwise comparisons to compare climate labels to each other. To account for multiple comparisons, we applied the Holm-Bonferroni correction27,28 and reported adjusted P values.

We conducted exploratory moderation analyses by adding an interaction term to the regression models to determine whether the effect of climate labels on primary outcomes varied by participant characteristics (age category, gender, race and ethnicity, political party affiliation, and climate concern) or experimental condition in the unrelated parent study. Statistical tests were 2-tailed and used a critical α of .05. All participants with complete outcome responses were included in the analysis. All statistical analyses were performed using Stata/SE, version 18.0 (StataCorp LLC).

Results

Of the 15 977 individuals assessed for eligibility, 2970 were screened out due to full quotas, 1538 purchased food from restaurants less than once per month, 605 quit before randomization, 227 failed an attention check, 166 did not complete eligibility questions, 81 completed the survey unrealistically quickly (less than one-third median completion time), 25 individuals were younger than 18 years, 19 did not select a menu item after randomization, 3 lived outside the US, and 1 responded twice so their second response was dropped. The final analytic sample included 10 342 participants (Figure 1).

Figure 1. CONSORT Flow Diagram.

Figure 1.

Open in a new tab

Among the 10 342 participants, the mean (SD) age was 46.4 (16.9) years; 53% were women, 47% were men, and less than 1% reported having another gender identity. Five percent of participants self-identified as non-Hispanic Asian; 13% as non-Hispanic Black or African American; 16% as Hispanic (any race); 1% as non-Hispanic Pacific Islander, American Indian, or Alaskan Native; 61% as non-Hispanic White; 3% as non-Hispanic multiple races; and less than 1% as an unlisted race (Table). Across conditions, participants most frequently selected beef burgers (48%), followed by chicken or fish items (35%), salads (13%), and plant-based burgers (4%) (eTable 2 in Supplement 2).

Table. Participant Characteristics (N = 10 342)a.

Characteristic No. (%)
Age category, y
18-34 3171 (30.7)
35-54 3554 (34.4)
≥55 3617 (35.0)
Age, mean (SD), y 46.4 (16.9)
Gender identity
Woman 5477 (53.0)
Man 4820 (46.6)
Another gender identity 45 (0.4)
Race and ethnicity
Asian, non-Hispanic 541 (5.2)
Black or African American, non-Hispanic 1347 (13.0)
Hispanic 1662 (16.1)
Pacific Islander, American Indian, and Alaskan Native, non-Hispanic 133 (1.3)
White, non-Hispanic 6328 (61.2)
Multiple races, non-Hispanic 316 (3.1)
Race not listed, non-Hispanic 8 (0.1)
Education
Less than some college 3733 (36.1)
Some college 3236 (31.3)
Bachelor’s degree or higher 3373 (32.6)
Household annual income, US dollars
≤$35 000 3503 (34.8)
$35 001-$65 000 2898 (28.8)
$65 001-$95 000 1669 (16.6)
≥$95 001 2010 (19.9)
Household size
1 2766 (27.5)
2 3306 (32.8)
3 1805 (17.9)
≥4 2193 (21.8)
Political affiliation
Conservative 2890 (28.7)
Moderate 3729 (37.0)
Liberal 2548 (25.3)
No opinion 913 (9.1)
Climate concern
Not at all 1208 (12.1)
A little bit 1421 (14.2)
Somewhat 2683 (26.8)
Quite a bit 2421 (24.1)
A great deal 2295 (22.9)
Region
Northeast 1911 (18.5)
Midwest 2203 (21.3)
South 4234 (40.9)
West 1991 (19.3)
US territory 3 (0.0)
Open in a new tab
a

Missing data due to incomplete responses ranged from 0.0% to 3.0% and were not included in the denominator.

Primary Outcomes

In the control condition, 920 of 1728 participants (53%) selected an item with high climate impact. Compared with participants in the control condition, 777 of 1724 participants (45%) in the high climate impact condition (adjusted P < .001), 787 of 1721 (46%) in the climate grade plus scale condition (adjusted P < .001), 817 of 1725 (47%) in the warning high climate impact condition (adjusted P = .01), and 825 of 1721 (48%) in the climate grade condition (adjusted P = .02) were significantly more likely to select an item with high climate impact. There was no difference between participants in the environmental cost condition (851 of 1725 [49%]) and those in the control condition (adjusted P = .22) (Figure 2; eTable 3 in Supplement 2). After adjusting for multiple comparisons, there were no significant differences among climate label conditions.

Figure 2. Bar Chart Showing Percentages of Participants Who Selected an Item With High Climate Impact by Climate Label Condition.

Figure 2.

Open in a new tab

P values are shown for conditions with a significant difference compared to the control. P < .05 indicates a significant difference after applying the Holm-Bonferroni correction for multiple comparisons. Error bars represent SEs.

In the control condition, participant selections had a mean GHGE of 18.8 (95% CI, 17.9-19.6) kg CO2e. Compared with the control condition, the mean GHGE from selected menu items significantly differed in the climate grade plus scale condition (16.3 [95% CI, 15.5-17.2] kg CO2e; adjusted P < .001) and high climate impact condition (16.9 [95% CI, 16.0-17.8] kg CO2e; adjusted P = .04), but did not differ in the environmental cost condition (17.6 [95% CI, 16.8-18.5] kg CO2e; adjusted P = .73), climate grade condition (17.7 [95% CI, 16.8-18.6] kg CO2e; adjusted P = .70), or warning high climate impact condition (17.4 [95% CI, 16.6-18.3] kg CO2e; adjusted P = .44). Compared with the control condition, significant differences were equivalent to 13.3% (climate grade plus scale) and 10.1% (high climate impact) relative reductions in GHGE. There were no significant differences among climate conditions (Figure 3; eTable 3 Supplement 2).

Figure 3. Bar Chart Showing Mean Greenhouse Gas Emissions (GHGE) of Meals Selected by Climate Label Condition.

Figure 3.

Open in a new tab

P values are shown for conditions with a significant difference compared to the control. P < .05 indicates a significant difference after applying the Holm-Bonferroni correction for multiple comparisons. Error bars represent SEs. kg CO2e indicates kilograms of carbon dioxide equivalent.

The effect of climate labels varied by race and ethnicity for both primary outcomes as reported in eTable 4 in Supplement 2. The impact of climate labels on selecting an item with high climate impact also varied by climate concern. Compared with participants with none or a little bit of climate concern, the climate grade (309 of 772 [40.0%]; P value for interaction = .008) and high climate impact (279 of 762 [36.6%]; P value for interaction = .006) labels were more efficacious for participants expressing quite a bit or a great deal of climate concern. Effects did not vary by age, gender, political party affiliation, or experimental condition in the unrelated parent study (eTable 4 in Supplement 2).

Secondary Outcomes

There were no significant differences in NPI score among label conditions (Figure 4; eTable 5 in Supplement 2). Participants in the climate grade plus scale condition were the only participants to select a menu item with significantly less saturated fat (11.8 [95% CI, 11.4-12.2] g; adjusted P = .04) than those in the control condition (12.6 [95% CI, 12.2-13.0] g). In addition, participants in the climate grade plus scale condition selected a menu item with significantly fewer calories (650 [95% CI, 638-663] kcal; adjusted P = .02) and less sodium (1292 [95% CI, 1270-1314] mg; adjusted P = .02) than participants in the high climate impact condition (mean calories: 680 [95% CI, 667-692] kcal; mean sodium: 1342 [95% CI, 1320-1364] mg) (Figure 4; eTable 5 in Supplement 2).

Figure 4. Bar Charts Showing Mean Values for Secondary Outcomes by Climate Label Condition.

Figure 4.

Open in a new tab

The panels show mean values for Nutrient Profile Index (NPI) score (A), energy (B), saturated fat (C), and sodium (D) by climate label condition. P < .05 indicates a significant difference after applying the Holm-Bonferroni correction for multiple comparisons. Error bars represent SEs; horizontal brackets, significant pairwise comparisons.

Compared with the control condition, participants in the climate grade condition (700 of 1640 [43%]; adjusted P < .001), climate scale plus grade condition (671 of 1645 [41%]; adjusted P < .001), high climate impact condition (689 of 1643 [42%]; adjusted P < .001), and warning high climate impact condition (647 of 1644 [39%]; adjusted P < .001) were significantly more likely to notice their label; participants in the environmental cost condition showed no difference from the control condition. Participants in environmental cost condition (775 of 1720 [45%]; adjusted P < .001), climate grade condition (858 of 1718 [50%]; adjusted P < .001), climate scale plus grade condition (859 of 1715 [50%]; adjusted P < .001), high climate impact condition (825 of 1720 [48%]; adjusted P < .001), and warning high climate impact condition (808 of 1717 [47%]; adjusted P < .001) were significantly less likely to support the label as a potential policy option than those in the control condition (996 of 1725 [58%]).

Participants in the environmental cost condition (mean score, 3.0 [95% CI, 2.9-3.1]; adjusted P < .001), climate grade condition (mean score, 2.7 [95% CI, 2.6-2.8]; adjusted P < .001), climate scale plus grade condition (mean score, 2.6 [95% CI, 2.6-2.7]; adjusted P < .001), high climate impact condition (mean score, 2.7 [95% CI, 2.6-2.7]; adjusted P < .001), and warning high climate impact condition (mean score, 2.6 [95% CI, 2.5-2.7]; adjusted P < .001) were significantly more likely to perceive their assigned label as discouraging of purchasing an item with high climate impact than participants in the control condition (mean score, 2.1 [95% CI, 2.0-2.1]). The climate grade plus scale (859 of 1715 [50%]; adjusted P = .03) and climate grade (858 of 1718 [50%]; adjusted P = .04) conditions received significantly more support than the environmental cost condition (775 of 1720 [45%]). However, participants perceived the environmental cost label (mean score, 3.0 [95% CI, 2.9-3.1]) as more discouraging than all other climate labels (eTable 6 in Supplement 2).

Discussion

In this online randomized clinical trial, we compared the effects of 5 climate labels (environmental cost, climate grade, climate grade plus scale, high climate impact, and warning high climate impact labels) with a control label and to each other on the climate impact (primary outcomes) and healthfulness (secondary outcomes) of fast food menu selections. Participants in 4 of the 5 climate label conditions, but not the environmental cost condition, were less likely to select an item with high climate impact than those in the control condition. However, only participants in the climate grade plus scale and high climate impact conditions selected an item that had significantly lower GHGE (ie, smaller climate impact) than those in the control condition. For the secondary nutrition outcomes, NPI scores (ie, nutritional quality) were not significantly different between conditions; however, participants in the climate grade plus scale label condition selected items with significantly less saturated fat than those in the control condition and less calories and sodium than those in the high climate impact condition. These findings indicate that various climate label designs may reduce climate impact, but climate grade plus scale labels may be stronger designs for reducing climate impact and improving nutrition at fast food restaurants.

Although this study did not find significant differences between the 2 grade label designs or between the 2 warning label designs after adjusting for multiple comparisons, the climate grade plus scale and high climate impact labels tended to perform slightly better than the other grade and warning label designs. For example, participants in both the climate grade plus scale and high climate impact conditions selected an item with significantly lower GHGE than participants in the control condition, while those in the other grade and warning conditions did not. For the grade label designs, this finding aligns with previous research on nutrition labels showing that additional context aids interpretation.29,30,31,32 The climate grade plus scale label also has precedence for implementation in real restaurant and retail settings because the Nutri-Score label with a scale has been adopted by multiple European Union countries,33 and the Green-score label (formerly Eco-score) with a scale has been voluntarily implemented by some European food brands.34 For the warning label designs, our results are contrary to prior research finding that marker words, like Warning, are associated with higher perceived effectiveness35 and are preferred by young adults.36 The word High could have acted as a marker word on its own in the absence of the word Warning in the high climate impact label, which could partially explain our results. The red color of the label could have also inherently conveyed warning. Further research is needed to determine the necessity of marker words on climate labels and how they may interact with color. Across all grade and warning labels, there were no significant differences in noticeability, support as a policy option, or perceived discouragement.

There were no significant differences between the novel environmental cost label and the control label on primary or secondary nutrition outcomes. Although the environmental cost label outperformed all other labels on perceived discouragement, all other climate labels outperformed the environmental cost label on both noticeability and policy support. This finding may be due to the numeric, rather than interpretive, nature of the environmental cost label. Although the environmental cost label translated GHGE into more intuitive monetary values, it did not interpret how high or low those values were, in contrast with the grade and warning labels. The environmental cost label’s low noticeability compared with the other climate labels may have been because that label was black and white, and the environmental cost text was a smaller font than the text on other labels. Prior research indicates that black and white labels and small font may be less noticeable,37 interpreted inconsistently,38 and less likely to influence behavioral outcomes,37 although this may also depend on background menu color.36 Because participants perceived the environmental cost label as effective, further research is needed on the effects of various environmental cost label designs.

The effects of the labels were consistent across most participant characteristics, including age, gender, and political party affiliation, which is similar to other climate label studies.11 However, we did find that labels were more efficacious for participants reporting higher climate concern and minoritized racial and ethnic identities. This result is consistent with prior research showing that minoritized populations report higher environmental motivations for their dietary choices.5 These findings highlight that the tested labels have the potential to reduce the environmental impact and improve nutrition of menu selections across diverse populations.

Similar to existing literature,11,12,39 the differences in climate impact and nutrition outcomes between labels were significant but modest in magnitude. Compared with the control condition, climate labels reduced GHGE by 1.9 to 2.5 kg CO2e (10.1%-13.3% relative reductions), a fraction of the 31 kg CO2e average per capita, per day of food-related GHGE in North America.22 The significant differences in nutrition outcomes included a 0.8 g reduction in saturated fat (climate grade plus scale vs control), and 50 mg and 30 kcal reductions in sodium and calories (climate grade plus scale vs high climate impact). These modest effects may be due to the menu used in this study, which was based on a real fast food chain with items that have limited ranges for GHGE and nutritional quality. In an online climate label study with a wider range of menu item,39 climate labels similarly prompted more sustainable food choices than a control label, but there were few nutritional differences. Field studies of climate labels in real food environments have mostly taken place in university or workplace cafeterias with a wide array of food and beverage options and label designs. These studies have had somewhat mixed results.40,41,42,43 However, these settings are substantively different than fast food restaurants, so field studies of menu labels in real restaurants are needed to fully understand the potential effectiveness of climate labels for promoting sustainable, healthy diets. Although effects in the current study were modest, small GHGE reductions at the individual level could have large population-level impacts. Future research is needed to determine the effect of climate labels in real fast food settings.

Strengths and Limitations

Strengths of this study include the experimental design; large sample size with similar sociodemographic characteristics to the US; realistic online fast food menu and labels designed to make participants feel like they were placing an order online, a popular method for ordering fast food44; the chance to receive an item selected, incentivizing realistic behavior; the use of an anonymous, confidential survey to reduce social desirability bias45; and the behavioral outcomes measured.

This study is not without limitations. The methods used to calculate GHGE of menu items resulted in estimated values rather than actual values; the menu was based on only 1 fast food chain; menu items had a narrow range of healthfulness and climate impact; participants selected just 1 menu item from a truncated menu without side items or beverages; and outcomes were based on hypothetical choices in an online survey, which may not represent typical meals or diets when real monetary costs and hunger are involved.

Conclusions

In this randomized clinical trial of 10 342 participants, environmental labels indicating the climate impact of menu items, particularly climate grade plus scale labels, reduced the environmental impact and improved the nutritional quality of fast food purchases among US adults. Further research on environmental cost label designs and real food choices in various, real restaurant settings is needed to determine the ideal design for labels to promote more sustainable and healthier food choices.

Supplement 1.

Trial Protocol

jamanetwopen-e2619650-s001.pdf (378.1KB, pdf)
Supplement 2.

eTable 1. Survey Item Wording

eTable 2. Menu Items Selected by Food Category and Label Condition

eTable 3. Tables of Main Label Effects on the Selection of High Climate Impact Item and Greenhouse Gas Emissions (GHGE) – Underlying Results of Figure 2 and Figure 3

eTable 4. The Effect of Climate Labels on Main Outcomes, 1) Selecting A High-Impact-Item And 2) Mean GHGE, Moderated by Participant Characteristics

eTable 5. Table of Main Label Effects on the Average Nutrient Profile Index (NPI), Energy (kcal), Saturated Fat (g), And Sodium (mg) of Menu Items Selected – Underlying Results of Figure 4

eTable 6. Table of Main Label Effects on the Noticing a Label, Supporting a Theoretical Labeling Policy, and Perceived Discouragement

eFigure. Example Menu With Climate Grade Plus Scale Labels on All Main Menu Items

eAppendix. Expanded Greenhouse Gas Emissions Methods

eReferences

jamanetwopen-e2619650-s002.pdf (623.3KB, pdf)
Supplement 3.

Data Sharing Statement

jamanetwopen-e2619650-s003.pdf (16.9KB, pdf)

References

  • 1.Turcotte-Tremblay AM, Fortier G, Bélanger RE, et al. Adolescents’ impairment due to climate anxiety is associated with self-efficacy and behavioral engagement: a cross-sectional analysis in Quebec (Canada). BMC Public Health. 2024;24(1):3009. doi: 10.1186/s12889-024-20333-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Latkin C, Dayton L, Bonneau H, et al. Perceived barriers to climate change activism behaviors in the United States among individuals highly concerned about climate change. J Prev (2022). 2023;44(4):389-407. doi: 10.1007/s10935-022-00704-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Vieira J, Castro SL, Souza AS. Psychological barriers moderate the attitude-behavior gap for climate change. PLoS One. 2023;18(7):e0287404. doi: 10.1371/journal.pone.0287404 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.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]
  • 5.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]
  • 6.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]
  • 7.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]
  • 8.Pan A, Sun Q, Bernstein AM, et al. Red meat consumption and mortality: results from 2 prospective cohort studies. Arch Intern Med. 2012;172(7):555-563. doi: 10.1001/archinternmed.2011.2287 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.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]
  • 10.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]
  • 11.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. doi: 10.1001/jamanetworkopen.2022.48320 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.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]
  • 13.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]
  • 14.Grummon AH, Reimold AE, Hall MG. Influence of the San Francisco, CA, sugar-sweetened beverage health warning on consumer reactions: implications for equity from a randomized experiment. J Acad Nutr Diet. 2022;122(2):363-370.e6. doi: 10.1016/j.jand.2021.07.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.de Adelhart Toorop R, Yates J, Watkins M, Bernard J, de Groot Ruiz A. Methodologies for true cost accounting in the food sector. Nat Food. 2021;2(9):655-663. doi: 10.1038/s43016-021-00364-z [DOI] [PubMed] [Google Scholar]
  • 16.Michalke A, Stein L, Fichtner R, Gaugler T, Stoll-Kleemann S. True cost accounting in agri-food networks: a German case study on informational campaigning and responsible implementation. Sustain Sci. 2022;17(6):2269-2285. doi: 10.1007/s11625-022-01105-2 [DOI] [Google Scholar]
  • 17.Monetisation factors for true pricing. True Price Foundation. 2023. Accessed November 1, 2024. https://mcusercontent.com/f0b0dd2cfed21257f4a6efd21/files/1242c4ea-9010-417f-d6f4-9bbde1efaa7f/Monetisation_Factors_for_True_Pricing_v3_0_0.pdf
  • 18.Gashler K. Study to unmask the true cost of food for NYS agencies. Cornell Chronicle. 2022. Accessed August 29, 2025. https://news.cornell.edu/stories/2022/09/study-unmask-true-cost-food-nys-agencies
  • 19.The true cost of food. Cornell College of Agriculture and Life Sciences. Accessed September 8, 2025. https://cals.cornell.edu/food-systems-global-change/true-cost-of-food
  • 20.True cost of food: measuring what matters to transform the U.S. food system. The Rockefeller Foundation. 2021. Accessed October 16, 2024. https://www.rockefellerfoundation.org/report/true-cost-of-food-measuring-what-matters-to-transform-the-u-s-food-system/
  • 21.Waite R, Vennard D, Pozzi G. Tracking progress toward the cool food pledge. September 24, 2019. Accessed April 22, 2025. https://www.wri.org/research/tracking-progress-toward-cool-food-pledge
  • 22.Waite R, Blondin S. Identifying cool food meals. June 27, 2022. Accessed May 19, 2025. https://www.wri.org/research/identifying-cool-food-meals
  • 23.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]
  • 24.Grummon AH, Goodman D, Jaacks LM, et al. Awareness of and reactions to health and environmental harms of red meat among parents in the United States. Public Health Nutr. 2022;25(4):893-903. doi: 10.1017/S1368980021003098 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.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]
  • 26.Talbot D, Mésidor M, Chiu Y, Simard M, Sirois C. An alternative perspective on the robust Poisson method for estimating risk or prevalence ratios. Epidemiology. 2023;34(1):1-7. doi: 10.1097/EDE.0000000000001544 [DOI] [PubMed] [Google Scholar]
  • 27.Holland BS, Copenhaver MD. Improved Bonferroni-type multiple testing procedures. Psychol Bull. 1988;104(1):145-149. doi: 10.1037/0033-2909.104.1.145 [DOI] [Google Scholar]
  • 28.Aickin M, Gensler H. Adjusting for multiple testing when reporting research results: the Bonferroni vs Holm methods. Am J Public Health. 1996;86(5):726-728. doi: 10.2105/AJPH.86.5.726 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.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;41:529-550. doi: 10.1146/annurev-nutr-111120-094932 [DOI] [PubMed] [Google Scholar]
  • 30.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]
  • 31.Grummon AH, Musicus AA, Moran AJ, Salvia MG, Rimm EB. Consumer reactions to positive and negative front-of-package food labels. Am J Prev Med. 2023;64(1):86-95. doi: 10.1016/j.amepre.2022.08.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Grummon AH, O’Sullivan K, Petimar J, et al. Nutrition info and other front-of-package labels and simulated food and beverage purchases: a randomized clinical trial. JAMA Netw Open. 2025;8(10):e2537389. doi: 10.1001/jamanetworkopen.2025.37389 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.The Nutri-Score: a science-based front-of-pack nutrition label. International Agency for Research on Cancer. World Health Organization. 2021. Accessed September 9, 2025. https://www.iarc.who.int/wp-content/uploads/2021/09/IARC_Evidence_Summary_Brief_2.pdf
  • 34.The Green-score: eco-friendly choices made easier. Colruyt Group. Accessed September 9, 2025. https://www.colruytgroup.com/en/conscious-consuming/green-score
  • 35.Grummon AH, Hall MG, Taillie LS, Brewer NT. How should sugar-sweetened beverage health warnings be designed? A randomized experiment. Prev Med. 2019;121:158-166. doi: 10.1016/j.ypmed.2019.02.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Falbe J, Montuclard A, Engelman A, Adler S, Roesler A. Developing sugar-sweetened beverage warning labels for young adults. Public Health Nutr. 2021;24(14):4765-4775. doi: 10.1017/S1368980021002287 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Lemmon B, Gil A, Musicus AA, Hall MG, Roberto CA, Falbe J. Optimal design of an added-sugar menu warning label: a randomized experiment. Am J Prev Med. 2025;69(6):108038. doi: 10.1016/j.amepre.2025.108038 [DOI] [PubMed] [Google Scholar]
  • 38.Carrero I, Valor C, Díaz E, Labajo V. Designed to be noticed: a reconceptualization of carbon food labels as warning labels. Sustainability. 2021;13(3):1581. doi: 10.3390/su13031581 [DOI] [Google Scholar]
  • 39.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]
  • 40.Brunner F, Kurz V, Bryngelsson D, Hedenus F. Carbon label at a university restaurant—label implementation and evaluation. Ecol Econ. 2018;146:658-667. doi: 10.1016/j.ecolecon.2017.12.012 [DOI] [Google Scholar]
  • 41.Wolfson JA, Altema-Johnson D, Yett A, et al. Climate change menu labels in a university cafeteria: effects on student’s diets, perceptions, and attitudes. Appetite. 2025;211:108001. doi: 10.1016/j.appet.2025.108001 [DOI] [PubMed] [Google Scholar]
  • 42.Jones AD, Hoey L, Leung CW, et al. Greenhouse gas emission menu icons decrease service of all menu items in residential dining halls at a large, midwestern public US university. J Acad Nutr Diet. 2026;126(2):156206. doi: 10.1016/j.jand.2025.09.008 [DOI] [PubMed] [Google Scholar]
  • 43.ElHaffar G, Cardoza PA, Fakhry F, Dubé L. Nudging consumers towards the most environmentally friendly warm dish: a field experiment applying traffic light label and single label interventions in a hospital cafeteria. Appetite. 2025;214:108054. doi: 10.1016/j.appet.2025.108054 [DOI] [PubMed] [Google Scholar]
  • 44.Bennett R, Gomez-Donoso C, Zorbas C, et al. Prevalence of online food delivery platforms, meal kit delivery, and online grocery use in five countries: an analysis of survey data from the 2022 International Food Policy Study. Int J Obes (Lond). 2025;49(7):1307-1316. doi: 10.1038/s41366-025-01771-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Ried L, Eckerd S, Kaufmann L. Social desirability bias in PSM surveys and behavioral experiments: considerations for design development and data collection. J Purchasing Supply Manage. 2022;28(1):100743. doi: 10.1016/j.pursup.2021.100743 [DOI] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplement 1.

Trial Protocol

jamanetwopen-e2619650-s001.pdf (378.1KB, pdf)
Supplement 2.

eTable 1. Survey Item Wording

eTable 2. Menu Items Selected by Food Category and Label Condition

eTable 3. Tables of Main Label Effects on the Selection of High Climate Impact Item and Greenhouse Gas Emissions (GHGE) – Underlying Results of Figure 2 and Figure 3

eTable 4. The Effect of Climate Labels on Main Outcomes, 1) Selecting A High-Impact-Item And 2) Mean GHGE, Moderated by Participant Characteristics

eTable 5. Table of Main Label Effects on the Average Nutrient Profile Index (NPI), Energy (kcal), Saturated Fat (g), And Sodium (mg) of Menu Items Selected – Underlying Results of Figure 4

eTable 6. Table of Main Label Effects on the Noticing a Label, Supporting a Theoretical Labeling Policy, and Perceived Discouragement

eFigure. Example Menu With Climate Grade Plus Scale Labels on All Main Menu Items

eAppendix. Expanded Greenhouse Gas Emissions Methods

eReferences

jamanetwopen-e2619650-s002.pdf (623.3KB, pdf)
Supplement 3.

Data Sharing Statement

jamanetwopen-e2619650-s003.pdf (16.9KB, pdf)

Articles from JAMA Network Open are provided here courtesy of American Medical Association

RESOURCES