AI-generated research brief — verify at source
Alternative Front-of-Package Labels Outperform FDA’s Proposed Nutrition Scheme in US Adults
In a nationally representative randomised controlled trial of 13,929 US adults, a multilabel warning scheme highlighting only high-nutrient-content items (Multi-High-In) reduced selection of unhealthy foods by 5.0 to 7.7 percentage points compared with all other label designs tested, including the FDA’s proposed format. A modified version of the FDA scheme incorporating red highlighting for high nutrient levels (Nutrition-Info-Red) also outperformed the agency’s proposed label on consumer understanding, improving correct identification of the healthiest product profile by 5.4 percentage points.
What Was Studied
This trial investigated how different front-of-package nutrition label (FOPL) designs affect consumer understanding, purchasing behaviour, and perceived healthfulness of food products, outcomes that are critical to evaluating the FDA’s 2025 proposal for a mandatory standardised FOPL in the United States. Specifically, the study compared the FDA’s proposed scheme—which displays low, medium, or high descriptors alongside percent Daily Values for saturated fat, sodium, and added sugars—against three alternative labelling approaches and a no-label control.
How It Was Studied
This was a five-arm online randomised controlled trial, conducted between April 1 and 25, 2025, using the Qualtrics platform. A national sample of 15,582 US adults aged 18 years and older was recruited to reflect US demographic distributions, with 13,929 retained in the analytical sample after exclusions. Participants were randomly assigned in equal proportions (1:1:1:1:1) to one of five labelling conditions: no-label control, the FDA’s proposed Nutrition-Info-%DV scheme, a Nutrition-Info-Red scheme (no percent Daily Value, with “High” highlighted in red), a single High-In label listing all nutrients present in high amounts, or a Multi-High-In scheme using separate labels for each high-amount nutrient. Participants were masked to the study objectives to reduce response bias.
What Was Observed
- Nutrition-Info-Red outperformed the FDA’s proposed label on identifying product healthfulness. Consumers assigned to Nutrition-Info-Red were 5.4 percentage points more likely to correctly identify the healthiest nutrient profile (95% CI 4.1–6.8; p<0.0001) and 3.3 percentage points more likely to correctly identify the least healthy profile (95% CI 1.8–4.9; p<0.0001) compared with those viewing the FDA’s Nutrition-Info-%DV label.
- Multi-High-In produced the largest reductions in selection of high-nutrient foods. Compared with all other label schemes, Multi-High-In reduced the proportion of participants choosing products high in nutrients of concern by 5.0 to 7.7 percentage points across comparisons, the strongest behavioural effect observed in the trial.
- Multi-High-In also accelerated nutrient profile identification by up to 31.6 seconds. Participants viewing the Multi-High-In scheme identified nutrient profiles between 4.5 and 31.6 seconds faster than those in other labelling conditions, suggesting markedly improved at-a-glance usability.
- Unhealthy products appeared healthier under Nutrition-Info-Red and Nutrition-Info-%DV than under Multi-High-In. When products were high in one nutrient but low in two others, both the FDA’s scheme and the red-highlighted variant generated higher perceived healthfulness ratings than the Multi-High-In design, pointing to a potential risk of misleading impressions under those formats.
Why This Matters
This study provides direct empirical evidence relevant to an active US regulatory decision, with the FDA currently advancing a mandatory front-of-package labelling rule. The findings suggest that the agency’s proposed design may be less effective than warning-only approaches in guiding consumer choices and reducing selection of less nutritious products. Given that food labelling policy affects population-scale dietary behaviour, even modest improvements in label design could have substantial public health implications. The results may also inform labelling debates in other countries considering similar mandatory FOPL frameworks.
How to Read This Result
While the trial is large, well-powered, and nationally representative, its online experimental design may not fully capture the complexity of real-world grocery shopping, where time pressure, store layout, and habitual purchasing patterns could attenuate the label effects observed here.
Limitations
The abstract does not explicitly report study limitations.