Ultra-processed foods aren’t all the same – researchers propose new way to tell the difference

Member Matters

Ultra-processed foods aren’t all the same – researchers propose new way to tell the difference

22 July 2025

Research Feature

At the NUTRITION 2025 conference (hosted by the American Society for Nutrition in Orlando, Florida, 31 May – 3 June), researchers introduced a new way to differentiate processed foods — a tool called WISEcode UPF™ (Wc-UPF).

The most common system used to classify food processing is called NOVA, which groups all foods into four categories based on how processed they are. But there’s a catch. The final category, NOVA 4, lumps a wide range of foods — from yoghurt to frozen meals, chips to breads — into a single “ultra-processed” bucket. That’s raised concerns among the food industry and many experts, who say this kind of all-in-one label oversimplifies things and misses important differences between food products.

How is Wc-UPF different?

Rather than relying on broad descriptive categories, Wc-UPF uses data from more than 100,000 commercial food products and over 5,000 individual ingredients. Each product is scored based on three components:

The number of processed ingredients, weighted based on processing and health risks.

The percentage energy (kilojoules) from added sugar.

A penalty for ingredients with significant health concerns.

Based on this, foods are sorted into five categories:

Minimal (scores 0–3)
Light (4–6)
Moderate (7–10)
Ultra (11–15)
Super-ultra (>15)

Where the NOVA system classifies around 95% of foods as ultra-processed, the Wc-UPF tool found a much broader spread:

20% Minimal
16% Light
23% Moderate
20% Ultra
23% Super-ultra

Testing this further, the researchers looked at nearly 10,000 everyday grocery items commonly ordered through DoorDash™ — products like yoghurt, tortillas, ready meals, condiments, and ice cream. NOVA classified almost all of them as ultra-processed. But Wc-UPF found major differences:

Only 12% of yoghurts were considered ultra or super-ultra.
But more than 70% of ice creams landed in the top processing categories.

This shows that not all processed foods are the same — something NOVA struggles to show.

A few things to keep in mind

While these early findings are interesting, it’s worth noting a few important points:

The real-life testing only covered seven grocery categories (bagels, condiments, dressings, ready meals, ice cream, tortillas & wraps, and yoghurt).
The study has not yet been peer-reviewed, or the full methodology and results presented.
The research was funded by WISEcode LLC, the company behind the Wc-UPF tool.

Still, presenting the work at a major scientific conference suggests it’s gaining traction — and may influence how food processing is evaluated in the future.

AFGC perspective

While the research provides insights into how processed foods might be further differentiated, it’s important to clarify that the AFGC does not support classification systems based solely on food processing levels — whether it’s the NOVA system or newer tools like Wc-UPF.

Food classification must be evidence-based, relevant to health outcomes, and practical for industry and consumers alike. Simply grouping foods by how processed they are, without considering the overall nutrition quality, context of the whole diet, or role in food safety, accessibility and affordability, risks oversimplifying complex issues.

While the direction of this research may help refine the science, much more work is needed to ensure these measures are valid, consistent, and useful in real-world settings.

Reference:

Black R, Berciano S, Sathar S, Forester S, Reyes E, Li Z, Burton-Freeman B, Drewnowski A. Ultra-Processed Foods Are Not All Alike: A Novel, Objective Approach to Differentiate Among Processed Foods Including Those Classified As NOVA 4 Current Developments in Nutrition, 2025.

Member Matters