Ultra-processed or over-simplified? Reframing the narrative around ultra-processed foods.
Ultra-processed foods are everywhere — lining supermarket aisles, filling our newsfeeds, and fuelling one of the most polarising nutrition debates of our time.
But not all the noise is helpful. To move the conversation forward, health professionals need a clearer, more balanced understanding of what the research really says, and take care to avoid slipping into oversimplified nutrition messages that can, at times, do as much (or more) harm as they do good.
The debate that won’t quit
Few nutrition topics today stir as much heated debate — or confusion — as UPFs. Headlines swing from “Ultra-processed foods have been linked to weight gain and early death” to “Ultra-processed foods might not be the real villain in our diets”, leaving consumers confused and health professionals caught somewhere in between.
While the concept of UPFs has real value in highlighting the industrialisation of our food supply and its potential impact on population health, the public conversation has veered away from assessing the evidence and is careening toward moral judgment. “If it comes in a packet, it’s bad for you” is an easy soundbite and a seemingly simple solution for health. But as clinicians and nutrition communicators, we need to dive a little deeper and do more than provide easy answers; first, we need to make sure that we are asking the right questions.
Whenever someone talks about UPFs, they’re usually referring to the NOVA food classification system. Developed by researchers at the University of São Paulo in Brazil. NOVA is the most widely used framework for defining food processing levels used in research. It groups foods according to both the extent and the purpose of processing. It does not consider the nutritional composition at all (which is how food groups in the Australian Dietary Guidelines are categorised).
NOVA distinguishes four main groups:
Group 1: Unprocessed or minimally processed foods
These are whole foods such as fruits, vegetables, grains, meat, fish, eggs, and milk. Processing methods include drying, freezing, fermenting, pasteurising, or roasting — techniques that preserve foods or make them safe to eat without adding new substances.
Group 2: Processed culinary ingredients
These are substances extracted or refined from Group 1 foods or from nature and used in cooking to season and prepare other foods. Examples include oils, butter, sugar, salt, and vinegar.
Group 3: Processed foods
These are products made by adding ingredients from Group 2 (such as salt, sugar, or oil) to Group 1 foods, mainly to increase shelf life or enhance flavour. Examples include canned vegetables, cheeses, and freshly baked breads.
Group 4: Ultra-processed foods
These are industrial formulations made mostly or entirely from substances derived from foods (e.g., starches, isolates, sugars, and fats), as well as cosmetic additives that improve flavour, colour, texture, or shelf life. They typically contain little or no whole food, are designed to be convenient, hyper-palatable, and ready-to-consume, and are often heavily marketed. Examples include soft drinks, packaged snacks, sweetened breakfast cereals, mass-produced breads, and pre-prepared frozen meals.
NOVA’s definition is incredibly vulnerable to interpretation, and the categories can be slippery. What exactly makes something “formulated” or “industrial”? Everything from soft drinks and confectionery to fortified breakfast cereals, protein yoghurts, plant-based milks, and wholegrain breads can land in the same category. On the other hand, a plain Greek yoghurt might sit in Group 3 — but add a sweetener or emulsifier, and it jumps to Group 4. That’s an enormous spread: nutritionally, functionally, and contextually.
Where does the evidence stand?
There’s no shortage of studies linking higher UPF intake to obesity, cardiovascular disease, cancer, depression, and all-cause mortality. These associations are remarkably consistent across populations and study designs and do really represent a very real relationship.
But correlation isn’t causation, and here is where the complexities lie. To date, there is very little direct mechanistic evidence demonstrating a clear causal pathway between food processing levels and adverse health outcomes. On top of this, consumers with higher UPF intake often have lower overall diet quality, lower socioeconomic status, and less access to fresh food. The actual level of food processing may well be part of the problem. Still, it is difficult to separate the effects of processing from the nutritional content of the foods themselves, especially when intake tends to cluster within broader patterns of inequality, access, and lifestyle.
The science of UFPs is still evolving. Most data come from observational studies that rely on self-reported food intakes, which are prone to error and bias. Further complicating the story, many structured food-frequency questionnaires used to measure dietary intake (and therefore assess how much UPFs are consumed) lack sufficient information to determine their processing level accurately. For example, where a pasta sauce is recorded, was it homemade, or did it come from a jar? That information is typically missing, meaning conclusions are likely resting on category errors (where foods have been grouped or interpreted in ways that don’t truly reflect their classification) or have been extrapolated from other foods within the same NOVA category, even if their nutritional profiles differ substantially. Intervention studies are few, short, and often focus on narrow outcomes. And to make it even more confusing, although NOVA is the most widely used framework, there is no universally accepted definition of UPFs — a foundational issue that also needs to be addressed.
As a profession, we need to acknowledge that not all UPFs are created equal, and oversimplifying them into a single vast research category only muddies the picture. Some foods in the UFP category are clearly detrimental to health — like sugar-sweetened beverages. Still, others contribute important nutrients or convenience without compromising diet quality — fortified breads, yoghurts, infant formula, and many breakfast cereals are all examples of foods that can fall into the UFP category, but are, on balance, beneficial to health. Conflating all UPFs into one category and using it as a "catch-all" to make broad health-related recommendations misses the point: it’s not going to be just how a food is processed that matters, but also what the food is made from, how it’s consumed, and in what context it fits within the broader dietary pattern.
Researchers are continuing to explore potential biological mechanisms that might link food processing levels more specifically with health outcomes — for example, hyperpalatability, effects of food additives, and changes in the food matrices of ultra-processed foods — but, for now, the evidence remains mixed. We’re still learning which components, combinations, or contexts truly drive risk, and how they apply at both the individual and population levels.
The inherent privilege and the problem with “just avoid UPFs”
The oft-quoted advice of “avoid ultra-processed foods” or “just eat real food” sounds simple, but in practice, it’s anything but. It assumes time, money, equipment, and cooking skills — luxuries not evenly distributed across the population. For households already stretched for time, there’s not enough convincing evidence to say that making homemade versions of certain foods, like baking bread instead of buying it or making cookies from scratch rather than purchasing them, will lead to meaningful differences in either nutritional quality or health outcomes.
For many households, convenience foods are a daily necessity and, used well, can help to make healthy eating possible. A packet of wraps, or a ready-made stir-fry or pasta sauce, can help families prepare a meal quickly and affordably. It’s not just time and money that need to be considered, either — a bowl of cereal with milk can be an easy option for someone with limited physical function. Nutritionally, the gap between a homemade sauce and a bottled version may be small, but the gap in accessibility is enormous. In these cases, UFPs can serve as helpful facilitators for the consumption of nutritious foods.
When UPFs are oversimplified and framed as inherently “bad,” it risks shaming those who rely on them. It diverts attention from systemic issues like food access, pricing, and policy. The moralisation of food has never led to healthier populations, but it can lead to guilt, confusion, stigma, and exclusion.
Lessons from history
The 1990s “low-fat” movement taught us a powerful lesson about the dangers of oversimplifying nutrition science. In trying to help people reduce saturated fat, public health messaging reduced a complex issue to a single instruction: “Eat less fat.” Food manufacturers responded by reformulating products to be lower in fat. Although the message was well-intentioned, the accidental consequence became an entire new industry of foods higher in refined carbohydrates and added sugars, leaving supermarkets full of “low-fat” products with little nutritional value. At the same time, consumers avoided nutritous foods like avocados and nuts because of their fat content, a message that stubbornly persists to this day. And an entire generation of consumers became more confused than ever about what a “healthy” diet really looks like.
The same risk applies today with UPFs. A sweeping message to “avoid UPFs” may inadvertently discourage foods that are convenient, affordable, and nutritionally adequate, at a time when the global population is already micronutrient-deficient, while doing little to reduce intake of truly nutrient-poor options. Worse, it shifts focus from systemic solutions — like improving food environments and access — to individual blame. Simplicity helps people act, but oversimplification distorts the evidence and widens inequities. The goal isn’t to make nutrition messages more complicated; it’s to make them more accurate, compassionate, and context-aware.
The level of food processing is just one piece of the puzzle. It should be considered one variable among many: potentially helpful in comparing like foods against like, but not as a stand-alone determinant of health. When it’s treated as an absolute, we lose sight of what really shapes health: dietary quality, diversity, and the socioeconomic factors that govern access, affordability, and agency.
Where to from here?
Despite all the noise, we do have an excellent idea of what foods support health, and these remain strikingly consistent. Across decades of research and populations, the strongest predictors of health are patterns, not processing levels.
As a population, most people need to eat:
⬆️ More fruits and vegetables.
⬆️ More legumes, beans, and pulses.
⬆️ More wholegrains.
⬆️ More nuts.
And of course, balanced portions of lean meats, oily fish, eggs, and dairy (or fortified alternatives) to round it all out.
These foods don’t have to be “perfectly unprocessed”; they need to form the backbone of a diet that’s nutrient-dense, diverse, and achievable. The goal isn’t purity, it’s practicality. When it comes to health outcomes, food processing levels may well play a role, but they need to be considered alongside nutritional quality, context, and consumption frequency as part of the bigger picture.
For researchers, the task ahead is refinement: more precise definitions, more evidence from clinical interventions to explore and expand our understanding into causal pathways, and a sharper focus on mechanisms that separate nutrient-poor from nutrient-appropriate processing. Any potential impacts of food processing on health outcomes also need to be weighed against the benefits it can bring to population health through improvements in food accessibility, shelf-life, and food safety.
For communicators and clinicians, it’s about embracing nuance and translating the evidence without slipping into moralism or absolutism. The goal should be to focus on what builds better diets, and not just what breaks them down. That’s not a headline-grabber, and that’s part of the challenge: we need journalists and communicators who stay true to the science and don't chase clickbait.
And for consumers, the focus should stay on dietary patterns — how the whole diet looks across a day, a week, a month and over time. Not the processing level of any single food or beverage.
The FOODiQ Takeaway
The question we need to be focused on isn’t “Are ultra-processed foods good or bad for us?”, it’s “How can we make healthy, sustainable diets accessible and realistic for everyone?”.
That shift in focus matters: it moves the conversation from blame to balance, and from individual choices to the systems that shape them.
Food and nutrition scientists should keep asking hard questions about the entire food environment — for UFPs, this will mean refining definitions, strengthening evidence around causal pathways, and teasing out where processing helps, where it harms, and how innovation can bridge the gap between the two.
Public health should keep its focus on equity.
And nutrition communicators, dietitians, and nutritionists should keep doing what they do best: interpreting the evidence and helping people make informed, practical choices that support their health in ways that genuinely fit into their lives.