What does the evidence say about ultra-processed food?
The association with poor health outcomes is reasonably consistent; why is much less settled, and that gap matters for what anyone should actually do.
What the term means. It comes from the NOVA classification, which groups foods by degree of processing rather than by nutrient content. Group 4 — ultra-processed — covers industrial formulations typically containing ingredients not used in home cooking: protein isolates, modified starches, emulsifiers, colourings and flavourings.
The definitional problem is real. NOVA groups mass-produced sliced bread, baked beans and supermarket hummus alongside confectionery and soft drinks. Critics argue the category is too heterogeneous to be a useful causal unit, and that it captures a marker of dietary pattern rather than a mechanism.
What the evidence shows:
Observational studies consistently associate high UPF intake with obesity, cardiovascular disease, type 2 diabetes and mortality. These are large and reasonably consistent across populations — but observational, and UPF intake correlates with income, education, smoking and physical activity, so confounding is a serious concern.
One controlled trial stands out. Kevin Hall's randomised crossover study fed participants ultra-processed and unprocessed diets matched for calories, sugar, fat, fibre and macronutrients, with unrestricted eating. On the UPF diet, participants ate around 500 calories more per day and gained weight. This is important because matching the nutrients removes the obvious explanation.
Proposed mechanisms: UPF is typically energy dense and soft, so it is eaten faster and delivers more calories before satiety signals arrive — eating rate has emerged as a strong candidate. Also lower fibre and protein, hyper-palatability, and possible effects of emulsifiers on the gut, which is speculative.
Where caution is warranted. The trial was small and short. And the advice implied by the category is not obviously actionable — telling people to avoid UPF, when it is cheap, convenient and shelf-stable, is difficult for many households.
The safer conclusion: eating more whole foods and more slowly has better support than the category itself.