Nutrition science produces few tidy arguments, and the dispute over ultra-processed foods is no exception. Into that dispute comes a new finding: roughly nine in ten research papers that challenged the accepted definition of ultra-processed foods had some connection to the food industry. The study behind the figure examined authorship and funding disclosures rather than the health effects of any particular food, an important distinction its authors emphasise. Even so, the result has implications for regulators considering front-of-pack labels and for shoppers trying to make sense of competing claims. This explainer follows the approach of our earlier health coverage: evidence first, conclusions second.
Mapping the Disagreement
Method matters here. The team identified papers whose central purpose was to dispute the definition, rather than papers that simply used or tested it, and then checked conflict-of-interest statements and author affiliations against industry sources. Most of the critical papers, they report, had at least one author connected to manufacturers of ultra-processed products or to organisations those manufacturers support. The minority of critical papers without such ties are significant too: they show that definitional questions are also raised independently. The study’s contribution is to map the landscape of the disagreement, not to referee every scientific point within it or to assess individual products.
Processing, Not Just Nutrients
Under the NOVA framework, what defines an ultra-processed food is not a single nutrient but the nature and purpose of its processing. Group four products are industrial formulations, often made from extracted ingredients and cosmetic additives designed for shelf life, convenience and strong palatability. The system was built for researchers studying whole diets, and the Harvard School of Public Health is among the institutions that explain it for general readers. Disagreement usually centres on edge cases and usefulness: whether processing itself, rather than salt, sugar, fat or total calories, is the right lens. The new study deliberately stayed out of that question and examined the debaters instead.
Transparency as the Middle Ground
There is also a well-documented playbook in other industries, from tobacco to energy, in which manufacturing doubt about definitions and measures delayed regulation even when underlying evidence was strong. Nutrition researchers who cite that history are not claiming the situations are identical, only that definitional disputes deserve scrutiny about who is making them and why. Industry groups respond that their scientists contribute legitimate expertise and that excluding them would weaken the debate. Both positions can be examined on evidence. What the new analysis adds is proportion: if most definitional criticism comes from connected sources, policymakers may reasonably ask how much weight that criticism should carry. See also another recent explainer for related context.
Why Definitions Drive Policy
For regulators, definitions are not academic. A label, tax or advertising rule has to draw a line somewhere, and the NOVA boundary between processed and ultra-processed is exactly where disputes concentrate. Food companies facing new rules have strong incentives to argue the line is unscientific, while public health agencies argue that imperfect categories can still guide better diets. The funding analysis enters that policy debate as evidence about evidence. It does not design a label, and it does not prove that any specific rule would improve health. It does suggest that consultation responses and journal criticism should be read with their interests attached, which most regulators already say they do.
Habits for Evidence-Aware Shopping
Consumers can also use the episode as a quick lesson in reading science news. Check whether a striking claim comes from a study, a commentary or a press release; look for a funding statement; and notice whether independent experts are quoted. Those habits matter far beyond this one topic, because nutrition debates recur with each new ingredient, diet or labelling proposal. The researchers behind the analysis say their aim was transparency, not censorship: industry science can be published and debated, but openly. On that limited point there is wide agreement, and it may be the most useful conclusion shoppers can take from a complicated scientific row.
Researchers who work on diet and health add one more caution. Population studies linking high consumption of ultra-processed products with poorer outcomes are a separate body of evidence from the definitional papers examined here, and this analysis neither strengthens nor weakens them directly. Its narrower message is about the sociology of a scientific argument: when criticism of a measurement tool comes overwhelmingly from one interested direction, the disagreement may be less balanced than a simple count of papers suggests. Journals, funders and universities will draw their own lessons about disclosure. For everyone else, the episode is a reminder to ask not only what a study says, but who is speaking.
