Impacts of warning labels on ultra-processed foods among Latino adults: A randomized trial

In a randomized trial of 4,107 US Latino adults, both identity and health warning labels on ultra-processed foods significantly improved product identification and reduced perceived healthfulness and purchase intentions compared to control labels, with identity warnings proving most effective at identifying the food category.

The Gist

Imagine you are walking down the aisle of a grocery store. You grab a box of colorful cereal or a bag of fruity snacks. They look healthy, right? Maybe they say "fat-free" or "zero sugar" on the front. But deep down, these are actually Ultra-Processed Foods (UPFs)—foods that have been heavily engineered in a factory with lots of additives, rather than being made from whole ingredients like real fruit or grains.

The problem is that most of us can't tell the difference just by looking. It's like trying to spot a wolf in sheep's clothing without knowing what the wolf looks like.

This study asked a simple question: If we put a big, bold warning sign on these foods, would people finally realize what they are eating?

The Experiment: A Taste Test of Labels

The researchers gathered over 4,000 Latino adults in the U.S. (a group that often faces higher risks for diet-related health issues) and put them in a virtual grocery store. They showed them four popular, ultra-processed snacks (like yogurt drinks, salty chips, and sugary cereals).

But here's the twist: The snacks had different "signs" on them, like traffic lights for food. The participants were randomly assigned to see one of three signs:

1. The "Barcode" Sign (Control): Just a standard barcode. This is what you see now. It tells you nothing about the food's quality.
2. The "Identity" Sign: A black box that simply says: "WARNING: Ultra-processed food." It's like a name tag that says, "I am a factory-made snack."
3. The "Health" Sign: A black box that says: "WARNING: Consuming ultra-processed food and drinks can cause weight gain, which increases the risk of obesity and type 2 diabetes." This is like a doctor's note attached to the snack.

What Happened?

The results were like a lightbulb moment for the shoppers.

• The "Name Tag" Won the Recognition Contest: When people saw the simple "Ultra-processed food" label, 70% correctly identified the snack as unhealthy. When they saw the long "Health Warning," 67% got it right. Without any warning (just the barcode), only 54% realized it was a processed food.

  • Analogy: Think of the "Identity" label as a bright yellow caution tape that says "CAUTION: CONSTRUCTION." You don't need to read a manual to know to stay away. The "Health" label was like reading a detailed safety report; it was also effective, but the simple "Caution" sign was slightly faster to catch the eye.

• Both Signs Made People Less Hungry: Whether it was the simple name tag or the detailed health warning, both made people think the food was less healthy and less likely to buy it compared to the barcode.

  • Analogy: It's like seeing a "Do Not Enter" sign on a door. Even if you don't know exactly why you can't go in, you feel a little less tempted to push the door open.

The Surprising Twist: Who Noticed the Signs?

The researchers wondered if the signs worked the same way for everyone. They found some interesting patterns:

• The "Health Warning" was a Superpower for Some: The long, detailed warning about diabetes and weight gain worked best for people with higher education levels and those with limited English proficiency.

  • Why? The researchers guess that people with limited English proficiency might have taken more time to carefully read the English text, absorbing the message more deeply. Meanwhile, those with higher education might have understood the complex health terms better.
  • Analogy: Imagine a complex map. A hiker with a compass (high education) can read the detailed map perfectly. A traveler who doesn't speak the local language (limited English) might stop and study the map intently because they have to, rather than glancing at it and walking past.

• The "Identity" Sign was the Universal Translator: The simple "Ultra-processed food" label worked well for everyone, regardless of education or language skills. It was the most straightforward way to say, "This isn't real food."

The Bottom Line

This study is like a dress rehearsal for a new rule that could change how we shop. It shows that warning labels work.

• If we want people to quickly spot the "wolf in sheep's clothing," a simple "Ultra-processed" label is the most effective tool.
• If we want to scare people away with specific health risks, a detailed health warning also works, especially if people take the time to read it.

The Takeaway: Just like we have warning labels on cigarette packs or speed limits on roads, putting a clear, bold warning on ultra-processed foods helps consumers make smarter choices. It doesn't just tell us what the food is; it helps us decide if we really want to eat it.

Technical Summary

1. Problem Statement

Ultra-processed foods (UPFs), defined by the NOVA classification system as Group 4 foods, constitute approximately 50% of food purchases and over half of daily caloric intake in the United States. High UPF consumption is strongly linked to chronic diseases, including obesity, type 2 diabetes, cardiovascular disease, and all-cause mortality.

Despite these risks, consumers often lack the ability to identify UPFs because the concept of "ultra-processing" involves complex industrial processes and ingredients not clearly conveyed on standard packaging. While front-of-package (FOP) warning labels have proven effective for nutrient-specific risks (e.g., high sugar or sodium), their efficacy in helping consumers identify UPFs remains largely unexplored. Furthermore, it is unclear which label design is most effective:

• Identity Warnings: Simply stating the product is "Ultra-processed."
• Health Warnings: Describing specific health consequences (e.g., weight gain, obesity, diabetes).

There is also a critical gap in understanding how these interventions perform among Latino adults, a demographic that faces higher rates of diet-related conditions and includes a significant portion of the population with Limited English Proficiency (LEP), who may face barriers to understanding text-based labels.

2. Methodology

Study Design:
An online, parallel-group, randomized controlled trial (RCT) conducted in August and September 2024.

Participants:

• Sample: 4,107 U.S. Latino adults aged 18–55.
• Recruitment: National convenience sample recruited via ThinkNow.
• Quotas: Approximately 50% of the sample had Limited English Proficiency (LEP), defined as speaking English "not at all" or "not well."
• Demographics: Mean age 35; 50% female; 49% LEP; 40% held a bachelor's degree or higher.

Intervention (Stimuli):
Participants viewed four real-world UPF products (yogurt, fruit drink, breakfast cereal, salty snacks) that were often perceived as healthy (e.g., "fat-free" or "zero sugar" claims). They were randomly assigned to one of three conditions (1:1:1 ratio):

1. Control: A barcode label (to control for the visual presence of a label).
2. Identity Warning: "WARNING: Ultra-processed food."
3. Health Warning: "WARNING: Consuming ultra-processed food and drinks can cause weight gain, which increases the risk of obesity and type 2 diabetes."

Measures:

• Primary Outcome: Correct identification of the product as ultra-processed (Yes/No/Not sure; "No" and "Not sure" combined).
• Secondary Outcomes:
  • Perceived healthfulness (1–5 scale).
  • Purchase intentions (1–5 scale).
  • Perceived Message Effectiveness (PME) in discouraging consumption (1–5 scale).

Statistical Analysis:

• Mixed-effects logistic regression for the primary outcome (accounting for repeated measures across four products).
• Mixed-effects linear regression for continuous secondary outcomes.
• Exploratory interaction analyses to test moderation by education level and English proficiency.
• Analyses conducted in Stata MP (v.18) with pre-registered hypotheses.

3. Key Contributions

• Novelty: This is one of the first studies to directly compare identity warnings vs. health warnings specifically for UPFs, rather than nutrient-specific warnings.
• Population Focus: It addresses a critical gap by focusing on Latino adults, a high-risk group for diet-related diseases, with a deliberate oversampling of LEP individuals to test label accessibility.
• Policy Relevance: The study provides empirical evidence to inform upcoming U.S. legislation (e.g., state bills and federal proposals in 2024–2025) regarding mandatory UPF labeling.

4. Results

Primary Outcome (Identification):

• Both warning types significantly improved correct identification of UPFs compared to the control (54%).
  • Identity Warning: 70% correct identification.
  • Health Warning: 67% correct identification.
  • Control: 54% correct identification.
• The Identity Warning was significantly more effective than the Health Warning ($p=0.007$) in helping participants identify the product as ultra-processed.

Secondary Outcomes (Perceptions & Intentions):

• Compared to the control, both warning types significantly:
  • Reduced perceived healthfulness ($p<0.001$).
  • Reduced purchase intentions ($p<0.001$).
  • Increased Perceived Message Effectiveness ($p<0.001$).
• Crucially, there were no significant differences between the Identity and Health warnings regarding healthfulness perceptions, purchase intentions, or PME. Both were equally effective at altering consumer sentiment.

Moderation Effects (Education & Language):

• Education: The Health Warning had a stronger impact on correct identification for participants with high education compared to those with low education.
• English Proficiency: Contrary to initial expectations, the Health Warning had a greater impact on participants with Limited English Proficiency (LEP) compared to those with high English proficiency.
  • Authors' Interpretation: This may be due to LEP participants having higher education levels on average in this sample, or because LEP participants spent more time reading the detailed health warning, increasing its effectiveness.

Product Variability:

• Results were consistent across product categories, though the effect was strongest for yogurt and weakest for pretzels.

5. Significance and Conclusions

Policy Implications:
The study demonstrates that FOP warning labels are a viable policy tool to help consumers identify UPFs. While Identity Warnings are slightly superior for the specific task of identification, Health Warnings are equally effective at reducing purchase intentions and perceived healthfulness. This suggests that policymakers could choose either format based on regulatory goals, or potentially use a combination.

Public Health Impact:
Given that UPFs drive chronic disease, these labels offer a mechanism to disrupt the "health halo" of products that appear healthy (e.g., low-fat yogurt) but are ultra-processed. The findings are particularly encouraging for the Latino population, showing that warnings can effectively reach and influence LEP subgroups.

Limitations & Future Directions:

• Hypothetical Setting: The study used an online survey with hypothetical purchase scenarios, not real-world shopping behavior.
• Exposure Time: Labels were viewed briefly; long-term exposure effects are unknown.
• Future Research: Real-world trials in grocery stores are needed to confirm if these changes in intention translate to actual sales reductions. Additionally, research should explore the "sweet spot" for the number of health effects listed on warnings to maximize comprehension across diverse literacy levels.

Conclusion:
UPF warning labels, whether focusing on identity or health consequences, successfully improve consumer recognition of ultra-processed foods and negatively impact their perceived healthfulness and purchase likelihood. These findings support the implementation of mandatory UPF labeling in the United States.