The concept of "ultra-processed foods" (UPFs) has rapidly gained prominence in discussions about diet and health, influencing media coverage, consumer guides, and even political policy proposals. However, a recent analysis of landmark studies suggests that the evidence linking UPFs to adverse health outcomes may be less definitive than widely assumed.
The definition of UPFs is often based on the presence of ingredients not typically found in home kitchens, such as high-fructose corn syrup or maltodextrin. The NOVA classification system, commonly used in UPF research, categorizes foods into four tiers, with the fourth tier encompassing these highly processed items.
Challenges in Defining and Studying UPFs
Nutrition scientists and experts have raised concerns about the clarity and consistency of the UPF definition. The distinction between normal processing and "ultra-processing" can be subjective, and some classifications have been criticized as overly broad. For example, wheat gluten, an ingredient used in some Asian cuisines and home cooking, is often classified as ultra-processed when added to packaged breads, leading to questions about whether its presence inherently signifies unhealthiness.
Much of the initial evidence linking UPFs to health issues like heart disease and cancer stemmed from observational studies. These studies, often based on diet surveys, can identify correlations but struggle to establish causation. They may not capture sufficient detail about the specific types of foods consumed, making it difficult to isolate the effects of processing itself.
Randomized Controlled Trials and Their Limitations
To address the limitations of observational studies, researchers have conducted randomized controlled trials (RCTs), considered a stronger method for determining cause and effect. Several prominent RCTs have indicated that UPFs contribute to negative health effects, such as increased calorie consumption and weight gain. These findings appeared to offer more robust evidence for the harms of UPFs.
However, a recent analysis published in the journal Science examined five key RCTs and found that the diets compared in these studies differed in ways beyond just the level of processing. The ultra-processed meal plans in these trials tended to be more calorie-dense and lower in fiber than their non-ultra-processed counterparts. These factors, independent of the processing level, can also lead to overeating and adverse health outcomes.
For instance, UPFs in some trials were noted to have softer textures, which can accelerate eating and contribute to consuming more calories. They also often had higher energy density and lower fiber content. According to Faidon Magkos, a professor at the University of Copenhagen and an author of the Science analysis, these characteristics are not universally present in all UPFs, nor are they exclusive to them. Many non-UPF foods share similar properties, and conversely, some UPFs can be harder in texture or higher in fiber.
The typical comparison in these trials has been between UPFs (NOVA group 4) and diets composed primarily of unprocessed foods (group 1). Critics argue that this extreme contrast makes it difficult to isolate the specific impact of ultra-processing itself, as opposed to the well-established benefits of whole foods or the known issues with traditional processed ingredients like refined flour and sugar.
Implications for Policy and Consumer Choice
Questions about the precise definition and demonstrable harms of UPFs have implications that extend beyond academic debate. Nutrition guidance based on imprecise or potentially flawed frameworks can lead to ineffective or counterproductive policies. For example, the focus on UPFs may inadvertently stigmatize healthier plant-based alternatives that are classified as ultra-processed. It could also lead consumers to make less healthy choices, such as opting for butter over a lower-saturated-fat vegetable oil spread, based on a misunderstanding of the UPF label.
Federal and state initiatives are considering incorporating the UPF concept into regulations for programs like school meals. Additionally, some programs are beginning to certify foods as "non-ultra-processed," which critics suggest may lend an unwarranted aura of health to conventionally unhealthy packaged foods. Experts suggest that a clearer understanding of what specific food characteristics drive negative health outcomes is needed to create more effective guidance and policies. A framework that cannot clearly identify the harmful aspects of the food system offers limited leverage against industries that profit from it.