Ultra-processed food: what the NOVA classification and BMJ 2024 actually show

What “ultra-processed food” actually means

The phrase “processed food” is used loosely. In nutritional epidemiology it has a precise meaning, set out in the NOVA classification developed by Carlos Monteiro and colleagues at the University of São Paulo in the 2010s. NOVA groups food not by nutrient content but by degree and purpose of industrial processing:

• Group 1 — minimally processed or unprocessed. Fresh, frozen, or dried whole foods: vegetables, fruit, legumes, grains, nuts and seeds, fresh meat and fish, eggs, milk, plain yoghurt, water.
• Group 2 — processed culinary ingredients. Items extracted from group 1 for cooking: olive oil, butter, salt, sugar.
• Group 3 — processed foods. Made from group 1 + group 2: bread baked from flour, yeast, water, and salt; canned vegetables in water; cheese made from milk and rennet; smoked fish.
• Group 4 — ultra-processed foods (UPF). Industrial formulations made mostly from substances extracted from foods (refined oils, refined sugars, modified starches, isolated proteins) combined with cosmetic additives — emulsifiers, stabilisers, colours, flavours, non-nutritive sweeteners — designed to be shelf-stable, hyperpalatable, and convenient.

A useful practical test for UPF is: read the ingredient list. If it contains five or more items, several of which are unfamiliar to a home cook (modified starch, lecithin, mono- and diglycerides, soy protein isolate, glucose syrup, flavourings, gums), it is almost always Group 4.

In Australia, UPF currently provides about 42% of energy in adults and around 47% in children, according to the most recent national nutrition surveys interpreted through NOVA. That figure has been rising for decades and is now broadly comparable to UPF intake in the United States and the United Kingdom.

A. Core clinical — what the evidence actually shows

The strongest contemporary evidence on UPF and health comes from three sources.

Randomised metabolic ward trial. The 2019 Hall et al. NIH metabolic ward trial randomly assigned 20 adults to two weeks of an ultra-processed diet, followed by two weeks of an unprocessed diet (or vice versa), matched for total energy availability, macronutrients, sugar, sodium, fibre, and presentation. Participants on the UPF diet ate approximately 500 kcal/day more and gained around 0.9 kg over two weeks; on the unprocessed diet they ate less and lost weight. This is the only trial of its kind, sample is small, but it is the strongest direct causal demonstration that UPF promotes excess energy intake when access and palatability are controlled.

Umbrella review of meta-analyses. The 2024 BMJ umbrella review by Lane et al. integrated 45 meta-analyses across 32 health outcomes from prospective cohort studies. Higher UPF consumption was associated with:

• Higher all-cause mortality (≈21% increased risk in highest vs lowest)
• Higher cardiovascular mortality (≈50% increased risk)
• Higher incident type 2 diabetes (≈40% increased risk)
• Higher incident overweight and obesity (≈55% increased risk)
• Higher incident depression (≈22% increased risk)
• Higher incident anxiety (≈48% increased risk)
• Higher common mental-health disorders overall

The evidence quality was rated convincing or highly suggestive for cardiovascular events, mortality, type 2 diabetes, depression, and anxiety. Other outcomes (specific cancers, asthma) had less consistent evidence.

Specific sub-categories. Within UPF, certain sub-types have the most consistent harm signals. Sugar-sweetened beverages have been linked across multiple cohorts to weight gain, type 2 diabetes, and cardiovascular events. Processed meats were classified by the International Agency for Research on Cancer (IARC) in 2015 as Group 1 carcinogens for colorectal cancer; red meat is Group 2A. Sugar-sweetened breakfast cereals, packaged biscuits, savoury packaged snacks, and confectionery sit close behind on the metabolic-and-cardiovascular signal.

Mechanisms. The proposed mechanisms — none yet definitively established as the dominant pathway — include:

• Energy density and hyperpalatability driving overeating
• Low fibre reducing satiety and adversely affecting glycaemic response and microbiome composition
• Rapid eating rate (UPF is engineered to be easy to chew) reducing satiety signalling
• Glycaemic variability from refined carbohydrates
• Specific additives (some emulsifiers, certain artificial sweeteners) modulating gut microbiota or inflammation
• Displacement effect — calories from UPF crowd out vegetables, legumes, fruit, fish, nuts, and whole grains

The mechanism research is active and not resolved. The clinical implications, however, do not hinge on mechanism — the consistency of the cohort signal plus the one metabolic ward trial is sufficient to recommend reducing UPF intake.

B. Evidence appraisal and controversies

“It’s just calories” critique. A subset of nutrition researchers argue that the UPF category is too heterogeneous to be useful and that the underlying problem is excess kilojoules, added sugar, saturated fat, and sodium — addressed by existing dietary guidelines. The 2019 Hall metabolic ward trial weakens this critique: even when total kilojoules, macronutrients, fibre, sugar, and sodium were matched on offer, the UPF diet led to greater intake and weight gain. Something about UPF as a category — not just its nutrient content — affects how much people eat.

Misclassification of NOVA groups. Studies vary in how they assign foods to NOVA categories — supermarket sourdough bread with added emulsifiers is classified differently in different studies. This adds noise. The effects are nonetheless consistent across reasonably-conducted cohorts, which suggests the signal survives misclassification.

Wholegrain UPF. Some packaged wholegrain breads, breakfast cereals, and plant-based meat alternatives have favourable nutrient profiles but are formally UPF by ingredient list. The honest answer is that within the UPF category, these probably sit at the less-harmful end — but few of the cohort studies have stratified that finely. For now, “minimally processed wholegrains, legumes, vegetables” is a more reliable target than “premium UPF substitutes”.

Industry-funded confounders. Some criticism of UPF research has come from industry-funded or industry-adjacent researchers. The independent peer-reviewed evidence is consistent enough that this is not a serious challenge to the broader signal.

Reverse causation. People who eat more UPF tend to have other risk factors (lower income, less time to cook, higher overall caloric intake). Cohort studies adjust for measured confounders; residual confounding is the central limitation. The metabolic ward trial mitigates this concern for the energy-intake/weight finding specifically.

Eating-disorder framing. As with intermittent fasting, language around UPF can shade into restrictive food rules and moralised eating. The clinically useful message is proportion, not purity. Encouraging people to build cooking skill, prioritise vegetables and legumes, and reduce highly processed snacks and beverages is constructive. Telling people to “never eat” specific categories invites restriction-and-rebound cycles and can trigger or worsen disordered eating.

The “real food” framing. Public discussion sometimes pitches UPF reduction as a return to a romanticised past of “real food”. The clinical evidence supports reducing UPF; the romanticised history does not need to be true for the evidence to hold. The constructive framing is forward-looking: build cooking confidence, lean on vegetables and legumes, shift the proportion.

C. Australian operations — what this looks like in general practice

The Australian general practice framing of UPF reduction is integrated into broader chronic-disease prevention rather than treated as a standalone intervention. It comes up during:

• A Heart Health Check (item 699) where dietary pattern is part of the cardiovascular risk discussion.
• A long consultation (item 36) or very long consultation (item 44) for weight, glucose, or chronic-disease management.
• A GP Management Plan (item 721) with Team Care Arrangement (item 723) when there is established type 2 diabetes, obesity-related comorbidity, cardiovascular disease, or chronic kidney disease. This combination opens access to up to five subsidised allied-health visits per calendar year, including with an accredited practising dietitian via Dietitians Australia.

(MBS / PBS items verified 2026-05-15 via WebSearch — workspace egress to mbsonline.gov.au + pbs.gov.au still blocked; spot-check confirms current.)

The current Australian Dietary Guidelines (2013) — and the Eat For Health consumer site — do not yet use the term “ultra-processed” prominently. They use “discretionary foods” — defined as foods high in saturated fat, added sugar, salt, or alcohol, and energy-dense without being nutritionally rich. The overlap with UPF is substantial. Australian adults currently get about 35% of energy from discretionary foods; the guideline target is under 35%, with most adults aiming for further reduction.

The 2025 update to the Australian Dietary Guidelines, currently in consultation through NHMRC and Eat For Health, is expected to incorporate UPF more explicitly. The Heart Foundation healthy eating resources and the Cancer Council Australia diet and cancer resources have both moved toward UPF-conscious framing in advance of the guideline update.

A practical Australian general practice order of operations:

1. Confirm the eating pattern. A brief 24-hour or 3-day food recall — usually with the dietitian where there is a GPMP — identifies the major sources of UPF in the diet. The top three in most Australians are sugar-sweetened beverages, packaged snacks (biscuits, savoury snacks, confectionery), and ready meals or takeaway.
2. Build the substitution list. Where the patient lands on UPF most often, what is the next-step minimally processed alternative they would actually use.
3. Cooking skill, not just food rules. The single best long-term intervention is cooking confidence with simple meals from raw ingredients.
4. Anchor changes to existing routines. Replace soft drink with water; replace packaged snacks with fruit, nuts, or vegetables and hummus; replace breakfast cereals high in sugar with rolled oats, plain yoghurt with fresh fruit, or eggs on wholegrain bread.
5. Track what matters. Weight, glucose, lipids, blood pressure, energy, mood — at the relevant clinical intervals.
6. Loop in a dietitian for complex cases — diabetes, kidney disease, food intolerance, family eating patterns, picky eating in children, cultural-food considerations.

D. Integrative considerations and a practical Australian framework

A reasonable, evidence-aligned target — and one consistent with the Heart Foundation and Australian Dietary Guidelines:

What	Target	Notes
Vegetables (incl. legumes)	5+ serves/day	Half the dinner plate; legumes 3–4 times/week
Fruit	2 serves/day	Whole fruit preferred over juice
Wholegrains	Most grain choices wholegrain	Rolled oats, wholemeal bread without long additive list, brown rice, barley, quinoa
Protein	Distributed across the day; mix of plant + animal	Fish 2–3 times/week; legumes regularly; eggs; modest red meat; minimal processed meat
Dairy or fortified alternative	2–3 serves/day	Plain milk, plain yoghurt; cheese in moderation
Olive oil as primary added fat	Daily	Replace butter or margarine in cooking
Water as primary drink	Most fluid intake	Tea and coffee are fine; limit sugar-sweetened beverages
Discretionary / UPF	Under 20% of energy ideally; under 35% as the floor	Sugar-sweetened beverages, packaged snacks, confectionery, processed meats

For practical Australian household-level guidance:

• The CSIRO Total Wellbeing Diet and Heart Foundation recipe library provide AU-tested meal patterns.
• The Eat For Health calculator translates Dietary Guideline serves into food-group targets for individual age and sex.
• Cancer Council Australia and Dietitians Australia maintain consumer-friendly summaries.

Where there are specific clinical conditions (chronic kidney disease, diabetes, food intolerances, eating-disorder history, pregnancy, paediatric feeding concerns), an accredited practising dietitian via the chronic-disease management pathway is the appropriate next step.

When to be cautious

A few situations where the simple “reduce UPF” message needs nuance:

• Eating-disorder history or current restrictive pattern. Frame around adding whole foods, building cooking skill, and inclusivity — not removing categories. Refer for specialist input if relevant.
• Tight budgets, food-insecurity contexts. Some UPF — bread, breakfast cereal, canned soups — is the affordable backbone of family eating. Practical guidance prioritises the highest-yield reductions (sugar-sweetened beverages, confectionery, packaged snacks) while leaving more functional UPF in place. Foodbank Australia and state-level food-relief services are part of the conversation where relevant.
• Children’s eating. Children’s UPF intake is higher than adults’ and has greater long-term implications. The conversation is family-level, not individual.
• Cultural-food contexts. Many traditional eating patterns are inherently low UPF and high in vegetables, legumes, and whole grains. Practical guidance leans on those rather than imported “wellness” food templates.
• Older adults at risk of undernutrition. The priority is total energy and protein adequacy; UPF reduction is secondary. Some convenience foods (oats, canned legumes, tinned fish) help meet protein and energy needs and should not be discouraged.

What this article is and is not

This is general health information drawn from current Australian Dietary Guidelines, peak-body resources (Heart Foundation, Cancer Council Australia, NHMRC, AIHW), and the major randomised and observational evidence on ultra-processed food and health. It is not personal medical advice and does not create a doctor–patient relationship. Decisions about individual dietary change are best made in the context of your medical history with your own GP and an accredited practising dietitian where appropriate.

For Australian consumer-friendly summaries: HealthDirect — Healthy eating, Better Health Channel — Healthy eating, and the Eat For Health Australian Dietary Guidelines portal.

---

Sources cited

1. NHMRC — Australian Dietary Guidelines
2. Eat For Health — consumer resources and 2025 update consultation
3. Heart Foundation — Healthy eating
4. Heart Foundation — Recipe library
5. Cancer Council Australia — Diet, nutrition and cancer
6. Therapeutic Guidelines (eTG)
7. Dietitians Australia
8. HealthDirect — Healthy eating
9. Better Health Channel — Healthy eating
10. AIHW — Nutrition across the life stages (2018)
11. CSIRO — Total Wellbeing Diet
12. Foodbank Australia
13. Monteiro et al. — NOVA classification (Public Health Nutr 2019)
14. Hall et al. — NIH metabolic ward UPF trial (Cell Metab 2019)
15. Lane et al. — UPF and health umbrella review (BMJ 2024)
16. Pagliai et al. — UPF and chronic disease meta-analysis (Br J Nutr 2021)
17. IARC — Red and processed meat carcinogenicity (Lancet Oncol 2015)