Scroll through health news for a week and you will meet the phrase ultra-processed food, usually attached to an alarming headline. Now a new study has examined the other side of that coverage — the papers arguing that the very definition of ultra-processed is flawed — and found that about nine in ten of them had authors with ties to the food industry. It is a finding about the debate itself, and a useful reminder to ask who is paying for the science we read.
To be fair to everyone involved: industry-linked researchers are not automatically wrong, and the classification at the heart of this argument has honest critics. Funding is context, not a knockout blow. But context matters, especially when a scientific disagreement doubles as a commercial one.
NOVA, in plain language
The definition comes from the NOVA classification, built by public health researchers in Brazil. Think of it as four shelves. Shelf one: foods as nature made them, or nearly — fruit, vegetables, eggs, plain milk. Shelf four: ultra-processed products, the industrial recipes — soft drinks, crisps, sweets, instant noodles, many ready meals — assembled from refined ingredients and additives you would not find in a home kitchen.
The argument is over shelf four. Critics say it sweeps too broadly, occasionally grouping foods with decent nutritional profiles alongside obvious junk. Supporters say the grouping captures something true about how these products are designed to be eaten quickly and in quantity. International guidance from the World Health Organization tends to favour diets based on the earlier shelves while scientists keep testing the details.
Why the funding pattern is news
Researchers who audit scientific literatures have seen this pattern before, in debates over tobacco, sugar and fossil fuels. Sponsorship rarely changes a result outright; it changes which studies get done, which get publicised, and how confidently conclusions are phrased. When nearly all the papers on one side of an argument share a funding source, the argument is not as open as it appears.
Practical readers can borrow the professionals’ habit: scroll to the conflicts section before believing the headline. Major nutrition schools, including the Harvard T.H. Chan School of Public Health, publish guidance that weighs hundreds of studies at once — a steadier basis for decisions than any single eye-catching paper, whoever funded it.
A bedtime reader’s guide to food headlines
Since this is a site for night owls, here is the late-night version of the method. When a food headline surprises you, ask four questions before sharing it. Who funded the study — the answer sits in the small print near the end. How many studies agree — one paper is a rumour, fifty are a literature. What exactly was measured — a definition, a diet, or a disease; the new study measured a definition debate. And what would change your mind — if nothing would, you are reading advocacy, not science.
Applied to this week’s story, the answers are reassuringly dull. The funding was lopsided, the finding fits a pattern seen in other industries, the subject was the classification argument rather than your dinner, and the correct response is curiosity rather than alarm. Save the alarm for the shopping itself: the evidence that mostly-whole-food diets serve people well remains exactly as strong this week as it was last week.
So what should shoppers do?
This study does not prove that every ultra-processed product damages health, and it would be wrong to treat it that way; even supporters of NOVA describe much of the outcome evidence as observational. What the study does justify is a bit more scepticism towards claims that the definition has been debunked, given who has financed most of the debunking.
The everyday advice remains boring and robust: build most meals from recognisable ingredients, treat shelf-four products as occasional rather than default, and read labels with the same suspicion you would bring to an advert. We made a similar point about evidence and patience in this recent article — good health information usually arrives slowly, in small print, and without a marketing budget. For more background reading, see this related explainer.
