What is an ultra-processed food?
“I know it when I see it.”
That quote comes from United States Supreme Court Justice Potter Stewart, who famously wrote those words in 1964 as he attempted to explain his definition of pornography.
The problem, as you might imagine, was not everyone saw what Justice Stewart saw. What some people thought of as pornographic and obscene, others described as artistic and beautiful.
For years, many people have used the “I know it when I see it” test with ultra-processed foods. However, as with the above example, this can also create confusion.
Take bread. Ask five clients whether it's ultra-processed, and you may get five different answers, ranging from “absolutely” to “no way” to “no clue” to “what’s ultra-processed?”
Partly because of this lack of clarity, the U.S. government reportedly may omit recommendations to cut back on ultra-processed foods in the upcoming revision of the US Dietary Guidelines for Americans.1
This is a stunning development given the growing body of research that has linked ultra-processed foods with poor health and early mortality.
When conducting studies about the health effects of ultra-processed foods, nutrition researchers often use the NOVA food classification system.
Developed several years ago by researchers at the University of São Paulo, Brazil, the NOVA continuum sorts foods into four categories based on their level of processing.2,3,4
However, the system is not without its flaws, as we explore in depth below.
What is the NOVA food classification system?
The NOVA system is a continuum that helps show the process of turning minimally-processed foods into ultra-processed ones.
The chart (below) shows the four NOVA categories, their definitions, and examples of where certain foods fall.2,3,4
To understand the weaknesses of NOVA, imagine what might happen if I printed out the above chart, handed it to you, and suggested you use it to consume fewer ultra-progressed foods. Based on the NOVA system:
How would you make decisions about flour-based foods like bread, pasta, and breakfast cereal?
What is more processed: a hot dog or a bowl of bran buds?
How would you decide which type of canned tuna or applesauce to purchase?
Would you choose store-bought hummus instead of Cheez-Whiz? Or would you believe the two options are nutritionally similar?
As you can see, there are a lot of gray areas.
This matters because some of the foods that fall into NOVA’s groups 2 through 4 seem to affect health very differently than others.
This brings us to the study I’d like to explore in today’s edition of Research Review. Published in the Lancet, it offers a lot of food for thought.
What the study found
Researchers used dietary questionnaires and medical records to track the eating habits and health outcomes of 311,892 people from eight European countries over nearly 11 years.
Then, researchers used the NOVA system to determine how food processing affected metabolic health.5
The overall finding wasn’t at all that surprising:
The study participants who consumed the most ultra-processed foods, according to the NOVA system, were also the most likely to develop type 2 diabetes throughout the 11-year study.
This finding aligns with past research linking ultra-processed foods with a range of poor health outcomes.6,7
However, when researchers dug into their data, they found that some ultra-processed foods affected metabolic health differently than others.
Notably, ultra-processed ready-to-eat meals, sugar-sweetened beverages, animal products, and savory snacks were associated with a higher risk of developing type 2 diabetes. However, plant-based alternatives, breads, and breakfast cereals were associated with a lower risk, as the chart below shows.
It makes sense that plant-based alternatives and bread products would result in a lower metabolic risk compared to other foods in NOVA category 4.
Soy products, for example, contain protein to slow digestion. Similarly, many breads and cereals contain fiber, along with nutrient fortification.
Is a soy sausage link a minimally processed food? Nope.
Is it less processed than chicken breast? Also, nope.
However, if you are moving along the continuum of a “little bit better,” I’d consider it a win if you swapped your usual pop tarts for a few soy sausage breakfast links.
However, according to the NOVA system, both the pop tarts and the soy links would fall into the same processing category.
Takeaways
1. NOVA isn’t perfect.
Because of how NOVA classifies foods, the Lancet researchers didn’t distinguish between white bread and whole-grain bread. Nor did they sort high-fiber cereal into a separate category from sugary cereal. Similarly, as mentioned above, soy burgers fall into the same nutritional category as donuts.
This illustrates some of the issues related to relying on NOVA.
However, NOVA falls short in other ways as well.
At last year’s Nutrition 2024 conference organized by the American Society of Nutrition, USDA scientists presented a paper that depicted two side-by-side food menus.
One menu was less processed than the other, according to NOVA’s classification. However, the less processed menu offered roughly the same nutritional value as the more processed menu—just at twice the cost.8
In the future, someone might develop a system that sorts foods more accurately than the current NOVA system.
But currently, NOVA’s four-category system is the best we have.
2. Other sorting systems also have problems.
Let’s explore a few other ways health professionals attempt to help you categorize foods, along with why they fall short.
▶ “Only shop the perimeter of the grocery store.”
The idea is that, by following this advice, you will automatically choose more minimally processed whole foods like fresh produce, dairy, and meat.
However, there are plenty of foods in the center of the grocery store that can help you move along the continuum of “a little bit better.”
If you never ventured into the center aisles, for example, you wouldn’t encounter frozen fruit or veggies, which are more convenient than fresh produce and sometimes contain more nutrients.
Same with dried or canned beans.
▶ “Avoid foods that come in a box or a bag.”
Like the above advice, this phrase means well.
However, if clients follow it carefully, they’ll avoid store-bought hummus and guacamole, canned beans, canned tomatoes, and dried whole grains.
▶ “Stick to foods with short ingredient lists.”
Depending on your definition of short, hummus will, yet again, likely fail this test.
(Not that anyone needs hummus in their diet, but it doesn’t deserve the exclusion it gets from this system of classification!)
3. People need help identifying ultra-processed foods.
I can’t assume you know which foods count as ultra-processed, which ones count as minimally processed, and which ones fall somewhere between.
It’s okay to use the NOVA system or some other visual as a starting point—I would just also want to discuss why certain ultra-processed foods are linked to poor health.
For example, I might review how added sugar, fat, and sodium affect a food’s nutritional quality. Or how the absence of certain nutrients—vitamins and minerals, phytochemicals, healthy fats, fiber, protein—changes a food’s nutritional profile.
With this understanding, you can make more informed decisions—and move along their continuum of “a little bit better.”
Closing Thoughts
The conversation around ultra-processed foods is nuanced, and tools like the NOVA classification system can help us begin to understand how food processing impacts health. However, as this study and others have shown, it’s not a perfect system. The NOVA framework offers a starting point, but it doesn’t account for the complexity of food choices, nor does it differentiate between foods within the same category that may have vastly different effects on metabolic health.
What matters most is the context of your diet as a whole. Instead of focusing on rigid classifications, consider the practical ways you can move along the continuum of "a little bit better." Small, meaningful changes—like swapping a sugary breakfast cereal for a higher-fiber option or choosing a plant-based alternative over a savory snack—can lead to better health outcomes without overcomplicating the process.
As we await advancements in food classification systems, it’s essential to keep perspective. Nutrition doesn’t have to be all-or-nothing. By understanding the limitations of frameworks like NOVA and focusing on balance, variety, and nutrient density, you can make informed choices that align with your goals, preferences, and lifestyle.
Remember, progress is better than perfection. Each meal is an opportunity to do a little bit better—not to be perfect, but to take another small step toward improved health. And in the end, that’s what matters most.
All the best,
Michael Beiter
Personal Trainer
Nutrition, Sleep, Stress Management, Recovery Coach
References
Cooney, Elizabeth. 2024. “Dietary Experts Advise Skipping Guidelines on Ultra-Processed Foods — for Now.” STAT. October 21, 2024. https://www.statnews.com/2024/10/21/ultra-processed-foods-more-data-needed-before-new-federal-dietary-guidelines/
Monteiro CA, Levy RB, Claro RM, Castro IR, Cannon G. A new classification of foods based on the extent and purpose of their processing. Cadernos Saude Publica. 2010; 26; 11: 2039-2049.
Monteiro CA, Levy RB, Claro RM, Castro IR, Cannon G. NOVA. The star shines bright. World Nutrition. January-March 2016; 7: 1-3: 28-38
Monteiro, Carlos A., Geoffrey Cannon, Renata B. Levy, Jean-Claude Moubarac, Maria Lc Louzada, Fernanda Rauber, Neha Khandpur, et al. 2019. Ultra-Processed Foods: What They Are and How to Identify Them. Public Health Nutrition 22 (5): 936–41.
Dicken, Samuel J., Christina C. Dahm, Daniel B. Ibsen, Anja Olsen, Anne Tjønneland, Mariem Louati-Hajji, Claire Cadeau, et al. 2024. Food Consumption by Degree of Food Processing and Risk of Type 2 Diabetes Mellitus: A Prospective Cohort Analysis of the European Prospective Investigation into Cancer and Nutrition (EPIC). The Lancet Regional Health. Europe 0 (101043): 101043.
Lane, Melissa M., Elizabeth Gamage, Shutong Du, Deborah N. Ashtree, Amelia J. McGuinness, Sarah Gauci, Phillip Baker, et al. 2024. Ultra-Processed Food Exposure and Adverse Health Outcomes: Umbrella Review of Epidemiological Meta-Analyses. BMJ 384 (February): e077310.
Vitale, Marilena, Giuseppina Costabile, Roberta Testa, Giovanna D’Abbronzo, Immacolata Cristina Nettore, Paolo Emidio Macchia, and Rosalba Giacco. 2024. Ultra-Processed Foods and Human Health: A Systematic Review and Meta-Analysis of Prospective Cohort Studies. Advances in Nutrition (Bethesda, Md.) 15 (1): 100121.
“NUTRITION 2024.” n.d. Accessed October 26, 2024. https://nutrition2024.eventscribe.net/index.asp?posterTarget=667035.
