P-0022 ● Active

“Ultra-processing itself causes people to overeat, beyond the nutrient content of the food.”

Community stance
No votes yet
§
Meaning & context

Diets high in ultra-processed foods (UPFs) are associated with obesity and many diseases. The question is whether processing itself is the problem, or whether UPFs are simply high in sugar, fat, salt and calories and low in fibre, in which case the label adds nothing beyond nutrient-based guidance.

Two small but carefully controlled feeding trials found that people eat more, or lose less weight, on ultra-processed diets even when nutrients are matched. Critics argue the NOVA classification is vague, that "nutrient matched" is never complete, and that specific properties like energy density and eating speed explain the effect better than processing as such. A 2025 Lancet series argued for UPF-focused policy; nutrition scientists remain split.

Ultra-processedContested
▶

NOVA group 4: industrial formulations made largely from substances extracted from foods or synthesised, with additives such as emulsifiers, flavours and colours (e.g. soft drinks, packaged snacks, mass-produced bread, reconstituted meats). The boundaries are disputed.

Beyond the nutrient contentContested
▶

Effects that remain after matching calories offered, macronutrients, sugar, salt and fibre: processing-related properties such as texture, energy density, eating rate or additives.

Filter by type:
↕
Arguments
6 total
↑ Supporting3
Causal

In a tightly controlled NIH inpatient trial, people offered ultra-processed meals ate about 500 more calories a day and gained weight, while the same people lost weight on minimally processed meals matched for presented calories, sugar, fat, fibre and salt.

Hall and colleagues randomised 20 adults to two weeks of each diet in a metabolic ward, eating as much as they liked. Intake was about 500 kcal/day higher on the ultra-processed diet, and participants gained about 0.9 kg versus losing about 0.9 kg on the unprocessed diet.

Hall et al. (2019). Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain. Cell Metabolism, 30(1), 67–77.

Limitation

Small sample; the diets differed in energy density and eating speed, which may explain the effect without invoking "processing". Two weeks is short.

Causal

A longer UK crossover trial found that a minimally processed diet produced roughly twice the weight loss of an ultra-processed diet, even though both followed national healthy-eating guidelines.

In the UPDATE trial, 55 adults ate each diet for eight weeks with food delivered at home and no intake restrictions; weight fell 2.06% on the minimally processed diet versus 1.05% on the ultra-processed one, with the extra loss coming from fat mass.

Dicken et al. (2025). Ultraprocessed or minimally processed diets following healthy dietary guidelines on weight and cardiometabolic health. Nature Medicine, 31.

Limitation

Both diets caused weight loss, so the ultra-processed diet that met guidelines was still healthier than participants' usual diets. Cardiometabolic markers barely differed.

Empirical

Higher UPF consumption is consistently associated with dozens of adverse outcomes across large cohorts, in ways that persist after adjusting for diet quality.

An umbrella review in the BMJ covering 45 meta-analyses and nearly 10 million participants found higher UPF exposure associated with 32 adverse health outcomes, with convincing evidence for cardiovascular mortality, type 2 diabetes and common mental disorders.

Lane et al. (2024). Ultra-processed food exposure and adverse health outcomes: umbrella review of epidemiological meta-analyses. BMJ, 384, e077310; Monteiro et al. (2025). Ultra-processed foods and human health: the main thesis and the evidence. The Lancet (Series).

Limitation

Observational associations cannot isolate processing from nutrients, income and lifestyle. Many associations are graded as weak evidence.

No arguments of this type on this side.
↓ Opposing3
Definitional

NOVA classification is too vague to be a causal variable. It lumps together wholemeal supermarket bread, flavoured yoghurt, soft drinks and processed meat, which have very different nutritional profiles and health effects.

Studies asking nutrition experts to classify foods using NOVA found poor agreement between them, especially for foods like bread, cheese and breakfast cereals.

Braesco et al. (2022). Ultra-processed foods: how functional is the NOVA system? European Journal of Clinical Nutrition, 76, 1245–1253; Gibney et al. (2017). American Journal of Clinical Nutrition, 106(3), 717–724.

Limitation

A fuzzy category can still track a real effect, and subgroup analyses (e.g. sugary drinks and processed meat drive most risk) refine rather than refute it.

Causal

Specific, measurable properties such as energy density, hyper-palatable combinations of fat, sugar and salt, and soft textures that speed eating explain overconsumption better than processing per se. Processed food engineered to lack these properties need not cause overeating.

Analyses of Hall's trial found that the ultra-processed diet had higher non-beverage energy density and was eaten faster. Fazzino and colleagues found that hyper-palatable foods, defined by nutrient combinations, made up most of the US food supply, overlapping heavily but not completely with UPFs.

Fazzino, Rohde & Sullivan (2019). Hyper-Palatable Foods: Development of a Quantitative Definition. Obesity, 27(11), 1761–1768.

Limitation

If processing reliably produces those properties, the distinction may matter little for policy.

Empirical

Some ultra-processed foods show no harm, or even lower disease risk: wholegrain breakfast cereals, packaged wholemeal bread and yoghurt. A causal effect of processing itself should not have such exceptions.

In the EPIC cohort (266,666 adults), the association of UPFs with cancer and cardiometabolic multimorbidity was driven by animal-based products and sweetened beverages; ultra-processed breads and cereals were not associated. In US cohorts, ultra-processed breads, cereals and yoghurt were associated with lower type 2 diabetes risk.

Cordova et al. (2023). Consumption of ultra-processed foods and risk of multimorbidity of cancer and cardiometabolic diseases. The Lancet Regional Health – Europe, 35, 100771; Chen et al. (2023). Diabetes Care, 46(7), 1335–1344.

Limitation

The exceptions may reflect nutrient content (fibre, whole grains) offsetting processing harms; this doesn't establish processing is harmless.

No arguments of this type on this side.
≈
Areas of agreement
3 points
Empirical

Diets high in ultra-processed foods are, on average, higher in calories, sugar, salt and fat and lower in fibre.

Empirical

Not all ultra-processed foods have the same health effects.

Empirical

The two key feeding trials are small and short.

P-0022 · Last edited Oct 26, 2026 · 1 revision