Illustrative stock photograph, not a Diva client, research participant or health result. No endorsement is implied. Photo: Gustavo Fring / Pexels.
The supermarket is an awkward place to resolve a scientific debate. You have a shopping list, a budget and perhaps twenty minutes before the next thing on your day’s schedule. A headline about ultra-processed food can make a familiar loaf of bread feel like a test you did not know you were taking.
The research deserves attention. Some controlled experiments show that the diets researchers provide can change how much people eat, and large observational studies raise concerns about long-term health. At the same time, “ultra-processed” covers a wide range of products. A useful answer needs more detail than a blanket instruction to empty the cupboard.
For adults trying to eat well, the practical question is what this evidence can reasonably change about everyday choices. Which findings come from experiments? What remains uncertain? And how do you use the information when fresh ingredients, cooking time, storage space and money are limited?
A sensible starting point is to keep two ideas in view: the way food is made can matter, and its nutritional contribution still matters. Neither the length of an ingredient list nor a fashionable front-of-pack promise can answer every question about a meal.
What the category actually describes
Much of this research uses NOVA, a system that groups foods according to the extent and purpose of processing. Its published explanation describes four groups: unprocessed or minimally processed foods, processed culinary ingredients, processed foods, and ultra-processed formulations.
That makes “processed” a much broader term than many conversations suggest. Freezing vegetables, pasteurising milk and milling grain are forms of processing. They do not automatically place a food in the ultra-processed group. Processing can improve safety, extend useful storage time and make food available when the fresh version is impractical.
The ultra-processed category generally involves industrial formulations containing substances or ingredients not commonly used in home cooking, often with additives serving sensory purposes such as flavour or texture. Familiar examples include many soft drinks, confectionery products and packaged snack foods. Some breads, breakfast cereals and other everyday products also qualify, depending on their formulation.
There are limits to what a shopper can confidently classify from a packet. A product name is not always enough. Two foods sold under the same broad name may have different ingredients and production methods. An unfamiliar ingredient name does not, by itself, establish harm, and counting ingredients is not a substitute for evaluating a product.
NOVA is also different from a nutrient profile. It does not directly tell you how much fibre, protein, salt or saturated fat a portion provides. That distinction helps explain why useful foods can fall within a category associated with poorer average outcomes.
The NHS guide to processed foods explicitly includes wholemeal sliced bread, higher-fibre breakfast cereals and baked beans among foods that can contribute to a healthy diet, including some classified as ultra-processed. A category can be useful for studying patterns while still requiring judgement about individual foods.
What a tightly controlled experiment found
One influential study moved the question beyond asking people what they remembered eating. In Kevin Hall and colleagues’ 2019 trial, twenty adults lived at a research centre for four weeks. They received an ultra-processed diet for two weeks and an unprocessed diet for two weeks, in random order, and could eat as much or as little as they wanted.
Participants ate about 508 more kilocalories a day, on average, during the ultra-processed diet. Their average weight increased by about 0.9 kilograms during that phase and decreased by about 0.9 kilograms during the unprocessed phase.
Random allocation and controlled food provision make this stronger evidence of cause and effect than a survey alone. The result applies to those tested diets under those conditions. It does not establish that every food in the ultra-processed group has the same effect.
The researchers tried to match several nutritional characteristics of the offered diets. Nevertheless, important differences remained, including the energy density of foods excluding beverages, and participants ate at different rates. The trial tested whole dietary packages, so it could not isolate one additive or one manufacturing step as the explanation.
It was also small and short. Living in a research centre with meals prepared removes the transport, cost, shopping and cooking decisions people make at home. The result is a valuable clue about food environments, but it cannot tell an individual exactly how their weight will change after replacing one item in a trolley.
A newer trial asked a more everyday question
A 2025 trial led by Samuel Dicken compared two diets designed to follow the UK’s Eatwell Guide. One relied on minimally processed foods and the other on ultra-processed foods. This is a useful question because people can improve nutritional quality without immediately removing every highly formulated product.
Fifty-five adults in England, with BMI from 25 to under 40 and habitual diets high in ultra-processed foods, were randomised. Fifty contributed to the intention-to-treat outcomes, and forty-three completed both diets. Each diet lasted eight weeks, with a four-week interval between them. Participants ate freely and received their food without charge.
Average body weight fell during both diets: by 2.06% during the minimally processed diet and 1.05% during the ultra-processed diet. The difference was about one percentage point. That finding supports the potential value of a minimally processed eating pattern, while also showing that the guideline-based ultra-processed diet was not associated with average weight gain in this trial.
Several details prevent a simple promise. The participants already ate diets high in ultra-processed foods, most were women, and the foods were supplied free. The diets were not identical apart from processing. Energy density was higher in the ultra-processed diet, adherence was partly self-reported, and the order of the diets and possible carryover effects complicate interpretation.
Mild adverse events were common during both diets, most often gastrointestinal symptoms. No related serious adverse events were reported. Neither menu can be treated as a personalised prescription for everyone with a digestive or medical condition.
Harvard Women’s Health Watch covered this study for a general audience. Its attention is understandable, but the original paper remains important: an eight-week average is not a forecast of what one person will lose over a year. It also does not demonstrate that a specific food caused a particular long-term disease.
Texture adds another piece to the picture
Research published in the April 2026 issue of the American Journal of Clinical Nutrition helps explain why the category itself cannot be the whole story. Forty-one adults completed two fourteen-day diets providing about 95% of energy from ultra-processed foods. The foods were selected to produce either slower or faster eating rates.
Average energy intake was about 369 kilocalories a day lower during the slower-eaten diet. There was no significant difference in body weight between the diets. Several dietary characteristics, including the energy density of foods excluding beverages, were matched.
This experiment suggests that physical properties affecting eating rate can matter within the same processing category. It does not prove that ultra-processed foods are harmless, and it does not prove that deliberately chewing any meal more slowly will reproduce the result. The researchers changed the foods they supplied, not simply the instructions given to diners.
It was another short study, with a relatively young group. Funding came through a public-private agri-food programme, and several authors disclosed food-industry relationships. Those disclosures belong alongside the methods and limitations when judging the evidence. They are not, by themselves, a reason either to dismiss the study or to treat it as the final answer.
Taken together, the experiments make a more interesting point than “one ingredient is the culprit”. Food structure, energy density, eating rate and overall nutritional composition can overlap. Sorting out their separate contributions takes careful study.
What the long-term headlines can and cannot tell us
Large observational studies can follow far more people for much longer than a tightly controlled feeding experiment. A 2024 BMJ umbrella review brought together pooled analyses linking higher ultra-processed food exposure with a range of adverse health outcomes.
These findings are a reason to take the subject seriously. They are not proof that a particular packaged food caused an individual illness.
People who eat different diets may also differ in income, smoking, activity, access to healthcare and many other ways. Researchers can adjust for measured differences, but some uncertainty remains. Dietary questionnaires are imperfect, and older records may not describe products well enough to classify them consistently.
It is also important to separate statistical strength from certainty about causation. A review can find a consistent association while judging the underlying evidence’s certainty to be low. A headline listing many diseases often loses that distinction.
The most useful response is neither panic nor dismissal. Long-term associations and shorter experiments provide different kinds of information. Together they justify paying attention to the overall pattern of eating, while leaving room for further research about mechanisms and differences between products.
They do not justify claims that all additives damage the gut, that a single convenience meal is toxic, or that avoiding this category guarantees protection from disease.
Use a label to answer a specific question
Trying to assign a perfect processing score to every purchase can make shopping exhausting. A more useful approach is to ask what you need that food to do.
If you are choosing a regular breakfast cereal, compare similar products. What does the nutrition panel say about fibre and sugars? If you are buying a soup or ready meal, how do the salt or sodium amounts compare? For everyday spreads or snacks, saturated fat may be another useful point of comparison.
Compare the same quantity, such as per 100 grams, where the labels provide it. Then consider the amount you actually eat. Different serving sizes can make one product appear lower in a nutrient simply because the suggested portion is smaller. Total sugars on a panel also do not always distinguish naturally occurring sugars from added sugars, so the ingredient list can provide additional context.
The aim is to compare realistic alternatives, rather than declare that the lowest number in every column must be best. A food’s contribution to the rest of the meal matters. Bread may provide a practical base for beans, eggs or another suitable topping. A convenient vegetable option may make the difference between including vegetables and going without them.
Front-of-pack terms such as “natural”, “plant-based” or “high-protein” do not settle these questions. A minimally processed product is not automatically suitable for every medical diet, and an ultra-processed classification does not erase a food’s useful nutrients.
Make the next shop easier, not more expensive
South Africa’s Guidelines for Healthy Eating and Food Guide place variety and practical food choices at the centre of advice. They also recognise the role of fortified maize meal and bread, particularly where diets lack variety. There is no good reason to discard an affordable staple solely because a broad online rule makes it sound inferior.
The World Health Organization’s healthy-diet guidance, updated in January 2026, emphasises adequacy, balance, moderation and diversity. Frozen and canned fruit and vegetables can be useful options when they do not contain added sugars or excessive sodium. Fresh-only rules can add expense and waste without answering the more important question of what you can reliably eat.
Start with one recurring situation. Perhaps breakfast is rushed, work lunches are unpredictable, or you arrive home too tired to cook. Choose a change that fits that situation and the facilities you have.
For breakfast, that could mean comparing two familiar cereals rather than buying an expensive specialist product. For an evening meal, it might mean keeping canned beans and vegetables available alongside a staple you already use. If you have freezer space, frozen vegetables can be a backup. If you do not, shelf-stable options may be more realistic.
Consider what the change adds as well as what it replaces. A better-stocked meal can be more useful than removing a convenient item and leaving a gap. There is little practical value in advice that produces a theoretically ideal shopping basket you cannot afford to refill.
Keep some flexibility for pleasure, culture and shared meals. One celebration or convenience meal does not describe an entire eating pattern. Nor is food processing a measure of effort, virtue or how much someone cares about their health.
If reducing a particular snack or drink feels useful, make a specific plan for the occasion when you normally have it. Identify an alternative you actually like and can obtain. An all-or-nothing rule offers little help when you are hungry at a petrol station or facing an unexpectedly long day.
Cost comparisons should include more than the shelf price. Storage, fuel or electricity, preparation time, transport and food waste all affect whether a choice works. The free meals supplied in research trials are a reminder that real households face constraints experiments may deliberately remove.
When general advice needs adapting
This article is written for adult education. People with medical dietary requirements may need different priorities. Kidney disease, food allergies, swallowing difficulties and some digestive conditions can make a generic shopping list unsuitable.
Prescribed nutritional drinks and specialised foods may be heavily formulated for a reason. Do not stop a clinically recommended product because it appears in an ultra-processed category. Ask the clinician or registered dietitian who knows why you use it about appropriate alternatives, if any.
If you have a history of an eating disorder, or food rules are becoming increasingly rigid, a processing-based exclusion list may make eating harder. Individual support can focus on adequate nourishment and a manageable relationship with food. Research about population averages should not become a reason to feel unsafe around every packet.
A useful professional conversation includes your budget, cooking facilities, preferences, symptoms and medication. Advice that ignores those details can sound precise while being difficult or inappropriate to follow.
Frequently asked questions
Is all packaged food ultra-processed?
No. Packaging is not the defining feature. Plain frozen vegetables, dried pulses and many other foods can be packaged without becoming ultra-processed. Classification depends on the food’s formulation and processing. The packet is also what keeps some foods safe and practical to store, so avoiding packaging is not a reliable nutrition strategy.
Should I stop eating sliced bread or breakfast cereal?
There is no universal requirement to remove either. Formulations vary, and some products in these groups can contribute fibre and other nutrients. Compare alternatives you can afford, consider what you eat with them, and keep medically necessary restrictions in mind. A useful everyday food does not become nutritionally empty because it fits one processing category.
Does a long ingredient list mean a product is dangerous?
No. Ingredient-list length alone cannot establish safety or nutritional quality. The experiments discussed here tested diets or food properties, not every individual additive. If you have an allergy, the relevant ingredient and allergen information is essential. That is a specific safety need, rather than a reason to fear all unfamiliar names.
Will cutting out ultra-processed foods guarantee weight loss?
No. The trial results are averages from particular diets and participants over limited periods. Your nutritional needs, starting eating pattern, portions, health and circumstances differ. Some people should not pursue weight loss at all. A change can be worthwhile for variety or nutritional quality without a promise about the scale.
Do I need to buy organic food or expensive “clean” brands?
No. Those labels do not establish that a product is minimally processed or that it meets your nutritional needs. Affordable staples, pulses and suitable canned or frozen foods can help build varied meals. Spending more is not a substitute for comparing the relevant information and choosing foods that fit your circumstances.
What is a reasonable first step this week?
Choose one frequently used food or one difficult mealtime. Compare a realistic alternative, or add a useful item that makes the meal easier to assemble. Keep the change small enough to repeat. If it raises cost, waste or stress without helping you eat adequately, adjust it rather than treating the difficulty as a personal failure.
A shopping basket is part of a life
Ultra-processed food research is developing quickly, and the strongest studies are helping to ask better questions. The evidence supports taking dietary patterns seriously. It also shows why the effects of food cannot be reduced to a single label.
You do not need to resolve every scientific uncertainty before your next shop. Keep useful staples, compare plausible alternatives, make room for variety and seek individual advice when medical needs change the picture. The goal is an eating pattern that supports your life and can be sustained within it.
Educational information for adults. This article is not a diagnosis, treatment plan or personalised nutrition prescription. Evidence and public guidance checked on 10 November 2026.