A fresh look at the research record behind the ultra-processed food debate has produced an uncomfortable statistic for the field. Roughly nine in ten papers that dispute how ultra-processed foods are defined were found to have ties to industry, according to the new study, which examined authorship and funding connections across the disputing literature rather than testing any diet itself.
Why would definition disputes attract such interest? Because definitions decide what regulation can see. If a labelling scheme, advertising rule or procurement standard is built on the ultra-processed category, then the boundary of that category determines which products are flagged and which are not. Challenging the definition’s validity can delay or reshape such policies without ever contesting the underlying health studies directly. Public health researchers have long noted the same dynamic in debates over tobacco, alcohol and sugar, where questioning a measurement can prove as effective as questioning a result.
It is equally important to state what industry ties do not show. Researchers with company connections can produce rigorous work, and some criticisms of the NOVA framework, such as borderline cases or inconsistent classification between coders, are raised by independent scientists too. Funding is a reason for transparency and scrutiny, not an automatic verdict on a paper’s methods. The balanced reading of the new study is that it documents an asymmetry in who argues one side of the debate, while leaving the scientific merits of each argument to be assessed paper by paper.
Counting the interests behind a definition
For labelling policy, the finding arrives at an active moment. Regulators in several countries are considering how processing should appear in dietary guidance or on packaging. Agencies such as the US Food and Drug Administration publish the current American labelling framework, the World Health Organization issues global nutrition guidance, and researchers at the Harvard T.H. Chan School of Public Health maintain accessible explanations of the evidence. None of that machinery moves quickly, and a single bibliographic study will not rewrite any label by itself this year.
For households, the practical takeaway remains modest and familiar. No single study, including this one, redraws sensible eating advice: build meals mainly from recognisable ingredients, treat heavily formulated products as occasional rather than default choices, and read labels for salt, sugar and fibre as well as marketing claims. The new research operates at a different level, examining the debate itself. Its nine-in-ten figure will likely be cited wherever ultra-processed labelling is proposed, and on the evidence presented, it deserves to be part of that conversation rather than a substitute for it.
Readers who follow our health coverage, including this recent report on the site, will recognise the wider pattern: nutrition science rarely moves in a straight line, and the way a study is framed can matter almost as much as its data when findings reach policymakers and the public.
A debate about a definition
Under the widely used NOVA system, processing itself is the organising principle. Canned vegetables and cheeses occupy middle groups, while ultra-processed foods are formulations assembled from ingredients rarely used in home kitchens, such as protein isolates, modified starches and flavourings. The definition matters because hundreds of studies have used NOVA categories to examine eating patterns, and because any future front-of-pack scheme based on processing would need a definition that inspectors, manufacturers and shoppers can all apply in the same way to the same product.
Against that background, the new study’s method was bibliographic rather than nutritional. It identified papers that disputed the ultra-processed definition and then examined their authors’ declared connections, funding sources and affiliations for links to food and drink companies or related organisations. The reported result, that about nine in ten such papers had industry ties, is a statement about the literature’s composition, not a measurement of any food’s effect on the body. That distinction is important, and the study’s framing keeps the two questions separate.
Studies of this kind are sometimes called research-on-research, because their subject is the scientific record itself. Here, the record under examination is the set of papers arguing that the ultra-processed definition is flawed, vague or unfit for policy. Checking those papers for industry ties required reading disclosure statements and funding declarations, which vary in completeness between journals. The headline proportion, around ninety percent, should therefore be read as the study reports it: a strong pattern in the disputing literature, identified through the connections placed on record.
