A new audit of research papers about ultra-processed foods offers readers something more useful than another diet headline: a lesson in how to read science. The researchers examined papers that disputed the standard definition of ultra-processed food and checked their authors for connections to the food industry. Nine in ten had such ties, including funding, consultancy, employment or affiliation with industry-linked groups. The audit does not accuse those authors of misconduct, and it does not claim to prove the definition right. It demonstrates, with unusual clarity, that the loudest printed scepticism about a public-health framework came predominantly from people connected to the industry that framework would regulate.
A revealing headcount
The framework is the NOVA classification, which sorts food by processing. Ultra-processed products, its fourth group, are industrial formulations: soft drinks, sweets, packaged snacks, instant noodles, reconstituted meats and similar items made from extracted ingredients and additives. Researchers use the category because population studies keep finding that diets dominated by these products are associated with higher calorie intake and greater incidence of chronic disease, even though such studies cannot by themselves prove processing is the causal ingredient. Critics respond that the grouping is crude, joining very different products, and that nutrition policy should focus on measurable nutrients. That is a real scientific argument, and it is stronger when made by researchers with no stake in the answer.
What NOVA is trying to capture
The audit’s value lies in proportions. Any single critical paper might be excellent; a literature in which nine of ten critical papers carry industry ties tells a different story about where criticism is manufactured and amplified. Similar audits in other fields changed how evidence is summarised, encouraging reviewers to ask whether conclusions survive when commercially linked work is removed. Public-health bodies already practise versions of this caution: the World Health Organization vets advisers for conflicts, and nutrition guidance from institutions like Harvard’s School of Public Health emphasises weighing the totality of independently produced evidence. The new study gives that caution fresh, quotable support in the specific argument over ultra-processed definitions.
Disclosure as a reader’s tool
For ordinary readers, disclosure works like an ingredients list for ideas. Just as a label lets you see what is in a product, a funding statement lets you see what is behind a claim. Making a habit of finding it takes seconds: scroll to the acknowledgements and competing-interest sections before sharing a headline. We suggested a similar discipline in this earlier report, where health claims also arrived with commercial baggage. The point is not to dismiss funded research unread; industry scientists do valid work, and some criticism of NOVA will improve it. The point is calibration: a claim aligned with its funder’s commercial interest should carry a burden of independent confirmation that a claim against interest does not.
Keeping the advice in proportion
Proportionality matters in the other direction too. This audit is a study of authorship, not a feeding trial; it cannot tell you what any food does to your body. It would be a mistake to treat it as proof that every ultra-processed product is dangerous or that the definition needs no refinement. Nutrition science is hard, diets are personal, and most researchers in the field, funded or not, agree on more than headlines suggest: meals built largely from whole and minimally processed ingredients, with adequate fibre and without excess sugar, salt and calories, are the consistent pattern associated with better health across very different research traditions.
The policy consequences will unfold slowly. Labelling rules and dietary guidelines are revised over years, through consultations where studies like this audit are cited by all sides. One plausible outcome is stricter, standardised disclosure, so that future audits are easy and routine rather than exceptional. Another is that journals will ask authors of definitional critiques to state industry relationships in the abstract, where readers actually see them. Either reform improves the information environment without silencing anyone, which is why transparency measures attract support across the scientific spectrum, including from industry researchers confident their work can withstand daylight.
Until those reforms arrive, the audit leaves readers better equipped rather than more confused. The definition of ultra-processed food remains debated, but the debate’s finances are now documented: the challenge has been financed largely by the challenged industry. Knowing that, you can read the next round of papers, headlines and label proposals with appropriately adjusted trust, giving independent replication the weight it deserves. In a media environment saturated with nutrition claims, that may be the healthiest takeaway of all: not a new rule about what to eat, but a sharper eye for who is telling you to eat it.
