Ultra-Processed Foods: The New Focus of Nutrition Policy

Ultra-Processed Foods

Chapter 2. What counts as ultra-processed food (and why definitions cause arguments)

If Chapter 1 was the “why now,” Chapter 2 is the “what exactly.” And here’s the uncomfortable truth: the argument over ultra-processed food is, first and foremost, an argument over definitions.

That may sound like academic hair-splitting. But realize that definitions are how research gets coded. They also determine how national surveys get interpreted. Additionally, they influence how headlines get written and how policies get enforced. Definitions decide which foods are targeted, which foods are exempt, which industries pay a cost, and which households feel blamed. They also shape the most intimate thing of all. Definitions affect how a person feels about feeding themselves. They also impact their emotions when they feed the people they love.

The term “ultra-processed foods” (UPFs) is most strongly associated with the NOVA classification system. NOVA does not classify foods by nutrients, but by the nature, extent, and purpose of processing. It introduced a distinctive claim: that a “level” of processing—especially industrial formulation with certain techniques and additives—might be independently relevant to health, beyond standard nutrients like fat, sugar, and salt (Monteiro et al., 2017; Monteiro et al., 2019).

But that’s exactly why people argue. Because “processing” is not a switch; it’s a spectrum. Because “industry” is not a moral category; it’s a toolchain that can preserve food safely or manufacture edible novelty. Because some foods that look wholesome are heavily formulated, while some foods that look “processed” are basically just preserved.

This chapter does four things:

1. It separates ordinary processing from “ultra-processing,” and shows where the line gets blurry in real life.
2. It explains NOVA Groups 1–4 with examples you’ll recognize.
3. It maps why experts disagree: category debates, reformulation battles, and “health halos.”
4. It ends with a calm, practical “spot-it-in-the-store” checklist that helps you navigate without fearmongering.

---

2.1. Processing vs. “ultra-processing”: where the line gets blurry in real life

Let’s start by rescuing the word processing from its internet reputation.

Processing, broadly, is anything you do to a food between farm/sea and mouth: washing, cutting, freezing, cooking, fermenting, drying, canning, pasteurizing, milling. Many of these processes exist because humans are mortal and bacteria are ambitious. Processing can improve food safety, reduce waste, extend seasonal availability, and make nutrient-dense foods easier to eat (think: frozen vegetables, canned fish, pasteurized milk, fermented dairy).

NOVA agrees with this in principle. It does not claim that all processing is harmful. It claims something narrower: that a specific pattern of industrial processing—fractioning whole foods into substances, modifying them, recombining them with additives, packaging them for convenience and hyper-palatable consistency—creates products that behave differently in diets and may drive health harms at population scale (Monteiro et al., 2019).

The core idea: “purpose of processing,” not “presence of processing”

NOVA is distinctive because it asks: Why was this food processed?

• To preserve and make edible? (freeze peas, can tomatoes, ferment yogurt)
• To create a stable, branded, ready-to-consume formulation designed for convenience and repeat purchase? (many packaged snacks, sugary drinks, ready meals, sweetened cereals)

Monteiro’s widely cited description of ultra-processed foods emphasizes industrial formulation, the fractioning of whole foods into substances, chemical modifications, assembly of ingredients “mostly of exclusive industrial use,” and frequent use of “cosmetic additives” (flavors, colors, emulsifiers, sweeteners), plus sophisticated packaging (Monteiro et al., 2019).

That’s the theory. The blur happens when you apply it to everyday products.

Where real life blurs the line (and why people get angry)

Blur #1: Bread, yogurt, and “ordinary” staples can land in Group 4

In Norway, bread is not an exotic indulgence; it’s breakfast, lunch, and sometimes a moral identity (“we eat grovt brød”). But many packaged breads can contain ingredients that push them toward NOVA Group 4 depending on formulation—emulsifiers, preservatives, enzymes, added fiber isolates, or other industrial ingredients.

This is not a “bread panic.” It is a classification dilemma: if a staple food becomes labeled “ultra-processed,” people understandably question whether the definition is capturing meaningful health risk or merely “how modern food is made.”

Nordic guidance recognizes this tension. The Nordic Nutrition Recommendations 2023 (NNR 2023) explicitly notes that some UPFs can be considered healthy from a nutritional point of view and that the category is heterogeneous; it also explains why the committee did not issue a specific UPF recommendation, arguing that the classification does not add clear value beyond existing food-based recommendations (NNR 2023).

Blur #2: Additives aren’t automatically villains—and “additive-free” isn’t automatically good

A product may contain a stabilizer to prevent separation, an antioxidant to prevent rancidity, or a preservative to reduce spoilage. Many additives are tightly regulated and used for functional reasons. Yet some people hear “additives” and think “toxic.” Meanwhile, “additive-free” can become a marketing halo even when the product is high in sugar or low in fiber.

UPF discourse often tries to walk a narrow ridge: it wants to critique the industrial formulation model without turning chemistry into a horror genre.

Blur #3: Reformulated products complicate everything

What if a “classic UPF” gets reformulated to reduce sugar, add fiber, or improve fat quality? Is it now “less ultra-processed”? NOVA, in its strict form, may still classify it as Group 4 if the industrial formulation and additive structure remain. Critics argue this can discourage beneficial innovation or mislead consumers by prioritizing processing category over nutrient quality (Gibney, 2018; Louie et al., 2025; Messina, 2025).

At the policy level, this matters because a lot of nutrition regulation over the last decade has focused on reformulation (reduce salt/sugar, add whole grains, etc.). Some authors argue that reformulation alone is not sufficient if the broader UPF pattern remains dominant (Scrinis et al., 2025).

Blur #4: Not all classification systems agree on what “highly processed” means

NOVA is the most famous, but it’s not the only way to define “highly processed.” A 2025 systematic review identified multiple food classification systems that include a “highly processed/UPF” category, underscoring that this is an evolving field rather than a settled dictionary entry (Medin et al., 2025).

This pluralism is part of the argument: if different systems draw different borders, then what exactly are we measuring when we say “UPF intake”?

What you should take from this section

• Processing is not inherently bad; ultra-processing is a specific pattern of industrial formulation and purpose.
• The border is blurry in real shopping life, especially for staples like bread and dairy.
• The blur isn’t a reason to ignore the topic; it’s a reason to read carefully and avoid purity thinking.

---

2.2. NOVA classification explained (Groups 1–4) and common examples readers recognize

NOVA is best understood as a map. It’s not the territory, but it tries to capture how foods function in diets and food systems.

NOVA Group 1: Unprocessed or minimally processed foods

These are edible parts of plants/animals/fungi/algae, plus minimal alterations that preserve or prepare without adding substances: washing, peeling, cutting, freezing, pasteurizing, drying, fermentation without major additions.

Examples (recognizable in Norway):

• Fresh or frozen vegetables and fruits
• Potatoes (raw), oats, rice
• Fresh fish, frozen fish fillets
• Eggs
• Plain milk, plain yogurt (depending on formulation)
• Coffee beans, tea leaves
• Nuts, seeds (unsalted/unsweetened)

NOVA Group 2: Processed culinary ingredients

These are substances extracted from Group 1 foods or from nature, used to cook and season: oils, butter, sugar, salt, starches. The idea is: you don’t typically eat these alone; you use them to make meals.

Examples:

• Olive oil, rapeseed oil
• Butter
• Sugar, honey
• Salt

NOVA Group 3: Processed foods

These are relatively simple products made by adding Group 2 ingredients to Group 1 foods, mainly for preservation or palatability. They usually have a short ingredient list and remain recognizable as the original food.

Examples:

• Canned tomatoes (tomatoes + salt)
• Salted nuts
• Cheese (milk + salt + cultures), many traditional cheeses
• Freshly baked bread made from flour, water, yeast, salt (depending on industrial additives)
• Canned fish (fish + oil or brine)

NOVA’s defenders often point out that Group 3 is “processed” but not “ultra-processed,” precisely because the food remains structurally recognizable and the formulation is relatively simple (Monteiro et al., 2019).

NOVA Group 4: Ultra-processed foods

This is the controversial one. NOVA describes these as industrial formulations made mostly from substances extracted from foods (oils, starches, proteins), derived from food constituents (hydrogenated fats, modified starches), or synthesized (flavors, colors, emulsifiers, sweeteners), typically with multiple ingredients and designed to be convenient, hyper-palatable, and heavily marketed (Monteiro et al., 2019).

Examples most people recognize immediately:

• Soft drinks and many sweetened beverages
• Candy, packaged desserts
• Chips, many packaged salty snacks
• Instant noodles
• Many ready meals (heat-and-eat trays)
• Many sweetened breakfast cereals
• Many packaged cookies, cakes, pastries

Examples that cause arguments because they feel “normal” or even “healthy”:

• Some packaged breads and crispbreads (depending on additives and formulation)
• Flavored yogurts, protein puddings, “diet” desserts
• Plant-based meat and dairy alternatives (often heavily formulated)
• “Fitness” bars, protein shakes
• “Gluten-free” and “free-from” specialty products
• Some fortified products where the base is highly refined

This is where the public debate often breaks: people don’t object to calling soda ultra-processed; they object to a model that can place a high-fiber packaged bread in the same Group 4 bucket as candy. Critics argue this lack of nuance can “demonize” nutrient-dense foods that contain certain additives, potentially distracting consumers from nutrient quality (Messina, 2025; Gibney, 2018).

Supporters respond that the category is meant to flag a food system pattern, not to claim every item in Group 4 is equally harmful.

A useful mental model: structure and substitution

When UPFs rise in a diet, what tends to fall? Often it’s the foods that require time, preparation, and perishability management—vegetables, legumes, fish, and whole foods that bring fiber and micronutrients.

The Nordic scoping review for NNR work notes that higher UPF intake tends to correlate with lower intake of fruits, vegetables, legumes, and seafood (Juul, 2024).

This “substitution effect” is one reason UPFs show strong associations with disease outcomes in observational studies: UPF-heavy diets can be both more of the harmful things and less of the protective things.

But the causal question remains: is it the UPF category itself (processing/formulation), or is it the nutrient profile and displacement of whole foods? Many expert bodies treat that as unresolved. For example, the UK SACN rapid evidence update concluded associations are concerning but emphasized uncertainty about whether harms are inherent to processing or driven by typical nutrient characteristics of many processed foods (SACN, 2025).

---

2.3. Why experts disagree: category debates, reformulation, and “health halos”

If you want to understand the “UPF wars,” don’t start with Instagram. Start with three real questions that scientists and policymakers genuinely disagree on.

Disagreement 1: Is NOVA valid enough for policy, or mainly a research lens?

NOVA has been enormously influential for monitoring diets and generating hypotheses about modern food environments (Monteiro et al., 2017). But critics argue it can be difficult to apply consistently (“food coding” problems), can misclassify foods, and may not align cleanly with nutrient-based risk models that have long guided policy.

A widely cited critique by Gibney argues that NOVA’s public health advice (“avoid UPFs”) overlaps heavily with known nutrient patterns (high sugars, low fiber) and raises questions about whether processing level adds independent predictive value beyond nutrients (Gibney, 2018).

Nordic authorities, notably, have chosen a cautious position: NNR 2023 recognized associations but decided not to issue specific UPF recommendations, citing heterogeneity and the view that the categorization does not add enough beyond existing classifications (NNR 2023).

This doesn’t mean NOVA is “wrong.” It means different institutions weigh trade-offs differently:

• Policy wants enforceable clarity.
• Science often produces probabilistic nuance.
• Public communication needs simplicity without distortion.

Disagreement 2: What about reformulation—does it meaningfully “solve” UPFs?

For years, much nutrition policy pressure has pushed reformulation: reduce salt, sugar, saturated fat; add fiber; fortify. The question is whether the UPF concern is mostly about these nutrient targets (in which case reformulation can help) or about the formulation model itself (in which case reformulation might not address the deeper issue).

A 2025 Lancet policy-focused piece argues that many policies have centered on shifting nutrients and promoting reformulation, but that UPF scrutiny demands broader evolution in policy thinking (Scrinis et al., 2025).

Meanwhile, UK SACN reflects the uncertainty: associations exist, but it remains unclear how much is due to processing per se versus nutritional characteristics (SACN, 2025).

In practical terms, reformulation creates political and consumer confusion:

• A cereal becomes “high fiber” and “reduced sugar” but remains an industrial formulation.
• A yogurt becomes “high protein” but contains sweeteners, stabilizers, and flavor systems.
• A plant-based burger has improved fat composition but is still made from isolates and additives.

Are these “better”? Often yes, nutritionally. Are they “not UPF”? Often no, by NOVA rules.

That clash is one reason the debate becomes heated: one side hears “UPF” as a warning against the industrial model; the other side hears it as an attack on food technology, affordability, and innovation.

Disagreement 3: The “health halo” problem—when labels and claims distort perception

The “health halo” effect is simple: when a product carries a positive cue (“wholegrain,” “high in fiber,” “natural,” “plant-based,” “protein”), people often perceive it as healthier overall and may underestimate sugar, calories, or degree of formulation.

This isn’t just a vibe; it’s measurable. Experimental research has shown that nutrient content claims can produce halo effects—raising perceived healthiness and intentions to buy, even when the underlying product is not meaningfully improved (Stoltze et al., 2021).

Packaging marketing also adapts to regulation and consumer expectations. A study documenting changes in marketing techniques and health/nutrition claims on product packaging (in a real retail environment) illustrates how quickly industry can shift messaging strategy (Saavedra-Garcia et al., 2022).

UPFs are particularly prone to health halos because their business model is compatible with claim-stacking:

• add a vitamin blend → “fortified”
• add chicory root fiber → “high fiber”
• swap sugar for sweeteners → “no added sugar”
• add protein isolate → “high protein”
• sprinkle seeds on top → “wholefood inspired”

Again: this does not mean the product is “evil.” It means the marketing surface can become more influential than the food itself.

The uncomfortable middle: both sides have a point

• NOVA highlights something real: the modern food environment is dominated by industrial formulations designed for convenience and high intake, and high UPF diets are consistently associated with poorer health outcomes across many studies (Juul, 2024; SACN, 2025).
• Critics highlight something real too: UPFs are heterogeneous, and a single category can mix foods with very different nutrient profiles and roles in diets—creating misclassification and confusion (Gibney, 2018; Louie et al., 2025; Messina, 2025).

This series will keep returning to a practical question that cuts through ideology:

Does this concept help people and institutions eat better—measurably—without increasing inequality or confusion?

Nordic institutions have been careful because the answer is not guaranteed.

---

2.4. A practical “spot-it-in-the-store” checklist

(ingredients, additives, packaging cues—without fearmongering)

You don’t need to memorize NOVA to shop well. You need a calm set of heuristics—tools that steer you toward better patterns while leaving room for real life.

This checklist is not meant to create “ingredient anxiety.” It’s meant to help you identify likely UPFs and decide when they’re worth it, when they’re not, and what to choose instead.

Step 1: Ask the simplest question first

“Could I reasonably make something like this in a normal kitchen, using normal ingredients?”

If the honest answer is “no,” it doesn’t automatically mean “don’t buy.” It means: treat it as a convenience product, not as the foundation of your diet.

Step 2: Ingredient-list signals (strong but not absolute)

Likely UPF signals:

• A long ingredient list with multiple unfamiliar substances
• Ingredients that look like industrial fractions rather than foods:
  • protein isolates, modified starches, maltodextrin, hydrogenated oils
• A cluster of cosmetic additives (terms vary by country):
  • flavors/aromas, colorants, emulsifiers, stabilizers, sweeteners
• Multiple forms of sugar (a common formulation trick):
  • glucose syrup, dextrose, fructose, invert sugar, etc.
• “Functional” additives that create certain textures:
  • humectants, gelling agents, thickeners, foaming agents

Usually-not-a-problem signals (context matters):

• A short ingredient list where you recognize most items
• One or two additives used for safety or basic function (not automatically a red flag)

NOVA’s formal descriptions emphasize that UPFs often rely on ingredients “mostly of exclusive industrial use,” plus additive systems to make products highly palatable and stable (Monteiro et al., 2019).

Step 3: Packaging cues (the “ecosystem” tells the truth)

UPFs often come with a familiar retail signature:

• Designed to be eaten anywhere (car, desk, stroller)
• Ready-to-eat or heat with minimal prep
• Shelf-stable for weeks/months
• Heavily branded with big claims on the front (“high protein,” “immune,” “fit”)
• Portion-optimized for snacking (single-serve packs, resealable pouches)

These cues don’t prove harm—but they strongly correlate with the UPF ecosystem.

Step 4: The “health halo” filter (don’t let claims do the thinking)

When you see front-of-pack claims, pause. Research shows that benefit-related nutrient claims can create a health halo, making products seem healthier overall and increasing intentions to choose them (Stoltze et al., 2021).

Try this:

• Ignore the front.
• Flip the package.
• Read ingredients + nutrition info.
• Then decide.

Step 5: A simple decision ladder you can use in 10 seconds

1. Base your basket on Group 1 foods when possible
   (veg, fruit, oats, eggs, fish, plain dairy, legumes, potatoes, whole grains).
2. Use Group 2 ingredients to cook more often than you think
   oil, salt, spices—because cooking is how you turn “ingredients” into meals that displace snacks.
3. Let Group 3 be your bridge category
   canned fish, canned tomatoes, cheese, simple breads—processed, yes; still “food.”
4. Treat Group 4 as optional—use it deliberately
   convenience for time-crunch days, travel, tight budgets, or pleasure—without letting it become the default.

Step 6: “Better UPF” swaps that avoid purity rules

If you’re buying a UPF anyway, you can still choose versions that align better with health patterns:

• Choose less sweet versions of yogurt/cereal/snacks
• Favor higher fiber and lower added sugar
• Prefer less salty snacks (and smaller portions)
• Pick ready meals with visible vegetables/legumes and lower saturated fat/salt when available
• Keep sugary drinks rare; choose water/unsweetened alternatives most days

This approach matches the cautious institutional stance seen in bodies like SACN: associations are concerning, but it’s not always clear whether harms come from processing itself or from common nutritional profiles of many processed foods (SACN, 2025).

The most important non-food part of the checklist

UPF intake is not just a shopping problem. It’s a time problem.

If your week is brutal, “just cook more” is not advice; it’s a guilt generator. The real skill is building systems:

• a few repeatable meals
• frozen vegetables and canned fish as tools
• one “batch cook” ritual if possible
• snacks that are actually foods (fruit + nuts; yogurt + berries; bread + egg)

UPFs win when life becomes too tight to manage meals. The goal is not perfection—it’s to keep the UPF share from quietly becoming the majority of your calories.

---

Where we land at the end of Chapter 2

• Processing is normal and often beneficial. “Ultra-processing” refers to a specific industrial formulation model and purpose (Monteiro et al., 2019).
• NOVA is useful but imperfect. It maps a real food-system pattern, but it can be hard to apply consistently and can mix nutritionally different foods in the same category (Gibney, 2018; NNR 2023; Messina, 2025).
• Experts disagree because the stakes are high: policy clarity vs. scientific nuance; reformulation vs. systemic change; consumer empowerment vs. confusion (Scrinis et al., 2025; SACN, 2025).
• You can shop intelligently without fear: treat UPFs as optional tools, watch for health halos, and build baskets around recognizable foods.

In Chapter 3, we’ll move from definitions to mechanisms: what could make ultra-processed foods uniquely problematic? We’ll examine satiety, texture, energy density, reward pathways, eating speed, food matrix effects, additives, and the “displacement” problem—always separating plausible mechanisms from proven causality.

---

References

• Gibney, M.J. (2018) ‘Ultra-Processed Foods: Definitions and Policy Issues’, Current Developments in Nutrition (PMC full text).
• Juul, F. (2024) ‘Ultra-processed foods – a scoping review for Nordic nutrition recommendations’, Food & Nutrition Research (PMC full text).
• Louie, J.C.Y. et al. (2025) ‘Are all ultra-processed foods bad? A critical review of the NOVA classification system’, Proceedings of the Nutrition Society.
• Medin, A.C. et al. (2025) ‘Definitions of ultra-processed foods beyond NOVA’, Food & Nutrition Research.
• Messina, M. (2025) ‘NOVA fails to appreciate the value of plant-based meat and dairy alternatives in the diet’, Nutrients (PMC full text).
• Monteiro, C.A. et al. (2017) ‘The UN Decade of Nutrition, the NOVA food classification and the trouble with ultra-processing’, Public Health Nutrition (PMC full text).
• Monteiro, C.A. et al. (2019) ‘Ultra-processed foods: what they are and how to identify them’, BMJ (PubMed record).
• Nordic Nutrition Recommendations (NNR) (2023) ‘Ultra-processed foods (UPFs)’, Nordic Council of Ministers (NNR 2023 web chapter / report).
• Saavedra-Garcia, L. et al. (2022) ‘Marketing techniques, health, and nutritional claims on packaging…’, Nutrients (PMC full text).
• SACN (Scientific Advisory Committee on Nutrition) (2025) Processed foods and health: Rapid evidence update (summary and report). UK Government.
• Scrinis, G. et al. (2025) ‘Policies to halt and reverse the rise in ultra-processed food…’, The Lancet (ScienceDirect record).
• Stoltze, F.M. et al. (2021) ‘Impact of warning labels on reducing health halo effects…’, Appetite (ScienceDirect record).

---

Optimizing Energy Performance with Digital Twins in Norwegian Buildings

---