Why is there so much mystification when it comes to food? Because the interests at stake are many, and frequently opposed. That is why, over the past few years, the debate on this most sensitive of subjects has taken on the tone of a full-blown culture war. On one side stand those who hold the food industry responsible for the rise in obesity, diabetes and metabolic disease. On the other are researchers, economists and industry professionals who urge us to distinguish between scientific evidence and ideological shorthand.
In this confrontation, few instruments have acquired the centrality of the NOVA classification, the system devised by Brazilian researchers that sorts foods into four categories according to their presumed degree of industrial processing. The most contested is the fourth: the "ultra-processed" foods, routinely cast as public health's enemy number one. The trouble is that the term has no scientific foundation whatsoever. So why does it strike such a chord?
NOVA's communicative power lies in its simplification: it takes on complex questions and answers them with a simplistic story. If a food is lightly processed, it is virtuous; if it is heavily transformed, it is suspect. But that very simplicity is the problem. The classification contradicts itself: oils, sugar and butter are treated as less processed than canned pulses or packaged bread, which may in fact be made with half the number of steps, or with gentler heat treatments.
The question is not whether nutritionally poor products exist. They do, and nobody disputes it. The question is whether the criterion of presumed "degree of processing" is really a scientifically valid tool for guiding consumer choice and, above all, for shaping public policy. The available evidence suggests the answer is far less obvious than we are usually told.
Link to A system even the experts struggle to applyA system even the experts struggle to apply
Among NOVA's many weaknesses, one stands out: it is remarkably hard to apply in practice. If a classification is to become the basis for dietary recommendations, taxation, labelling or trade restrictions, it ought to be straightforward to use and to produce consistent results regardless of who is using it. That is not what happens with NOVA.
A study published in the European Journal of Clinical Nutrition in 2022 by Véronique Braesco and colleagues asked some 200 food and nutrition professionals to classify a range of products using the NOVA system. The results showed extremely high levels of disagreement. The same foods were placed in different categories almost systematically, depending on who was doing the classifying.
"The system's consistency, functionality and coherence," the researchers write, "remain poorly characterised. Because its classificatory approach is purely descriptive in nature, it opens the door to ambiguity and to differing interpretations. Indeed, even experts run into difficulty and disagree when using it."
In other words, trained and competent professionals could not agree on how to classify the same products. This is an enormous problem. If a methodology yields different results depending on who applies it, it can hardly serve as a reliable basis for regulatory decisions affecting millions of citizens and billions of euros of market value. A classification that its own users cannot interpret consistently risks becoming an arbitrary instrument.
Link to The paradox of NOVA 4The paradox of NOVA 4
The second weakness concerns the internal architecture of the classification. Category NOVA 4 gathers wildly different products under the single label of "ultra-processed". Canned tuna in brine, flavoured infusions, fruit ice lollies, soft drinks, wholemeal bread, cured meats, snacks and tinned pulses can all sit side by side. In nutritional terms, in caloric terms, in functional terms and in terms of the role they play in the diet, these products have very little in common.
Some are high in protein and low in calories. Some deliver fibre, vitamins or essential nutrients. Others can be eaten occasionally with no particular consequence for health. And yet NOVA tends to treat them all as manifestations of the same phenomenon. It is rather like assigning a bank, a semiconductor plant and a hotel chain to the same economic sector simply because they all use computers. The category becomes so broad that it loses any analytical meaning. And when a classification is that heterogeneous, the obvious risk is that it ends up explaining little or nothing at all.
Link to When the process matters more than the nutrientsWhen the process matters more than the nutrients
The third weakness is perhaps the most serious. In the words of NOVA's own architect, "the issue is not food, nor nutrients, so much as processing". The classification therefore assesses a food's level of industrial transformation — and does so improperly at that — rather than its nutritional composition. As nutrition scientists and food technologists have pointed out, this leads to paradoxical outcomes. Tinned pulses and plant-based burgers can be classified as NOVA 4 despite being recommended sources of protein and fibre. Frozen vegetables can fall into the same category even though they often retain nutritional quality equal or superior to fresh produce that has been sitting around for days.
Some packaged wholemeal breads are labelled ultra-processed even though their nutritional profile is better than that of artisan loaves belonging to categories deemed less problematic. The upshot is that consumers are nudged into judging the production method — and judging it wrongly — rather than what is actually in the product.
This is a significant conceptual inversion, and a potentially dangerous one. Dietary and nutritional science has been refined over many decades through the study of measurable variables: calories, protein, fibre, fats, micronutrients and overall diet quality. Not opinions. NOVA shifts the focus onto the industrial process, implicitly assuming that transformation is in itself a marker of risk — while proving unable to define its level, its number or its intensity. It is, to put it mildly, a contentious assumption, and one that remains the subject of wide scientific debate.
Link to A battle about far more than foodA battle about far more than food
Behind NOVA's popularity lies a broader phenomenon. As many international observers have noted, the debate over ultra-processed foods has become a battleground for economic interests, regulatory strategies and ideological visions of society.
The food industry certainly has interests to defend. But this is a wide playing field, with numerous and competing stakeholders. It is no accident that activist organisations, academic groups and international bodies build consensus, visibility and influence around particular narratives.
The risk, in this context, is that the complexity of nutritional science is sacrificed to messages that are simple and easy to broadcast. Economic history teaches us that public policy works when it rests on criteria that are measurable, verifiable and coherent. When it is built instead on vague concepts and ambiguous or contested categories, it tends to produce unintended side effects.
A 2025 paper published by Cambridge University Press, by the researcher Jimmy Chun Yu Louie, put it this way:
"The binary nature of the NOVA classification takes no account of nutritional composition, of the benefits of fortification, or of cultural food traditions, generating inconsistencies in categorisation across different settings. The methodological limitations of UPF research include inadequate dietary assessment tools, selective reporting of negative findings, and design flaws that confound processing with other dietary factors. The difficulties of application extend to socioeconomic accessibility — since UPFs often supply low-cost nutrients to disadvantaged groups — and to environmental sustainability, a field in which indiscriminate reduction could increase resource requirements."
Link to Public health deserves better toolsPublic health deserves better tools
What too few people say with sufficient force is that the fight against obesity and metabolic disease is a genuine priority, and that our health, in the broadest sense, deserves better tools and better indicators than the ones currently in circulation. Precisely for that reason, it demands rigour.
The NOVA classification deserves credit for drawing attention to the relationship between modern diets and health. But its methodological weaknesses, its poor reproducibility, the heterogeneity of its categories and its limited capacity to reflect the real nutritional value of foods make it a problematic foundation for public policy.
To reduce the judgement of a food to its presumed degree of processing is to oversimplify a complex reality. Nutrition is not a battle between "natural" food and "industrial" food. It is the search for a balance grounded in evidence.
(Associated Medias) - Tutti i diritti sono riservati