UPF Brain & Cognitive Claims — Evidence Vault.
Source-validated peer-reviewed evidence for the thesis that ultra-processed food is engineered in ways that affect the brain — cognitive function, structural change, behavioural reward circuitry.
Stale-date reminder: re-check October 2026 — this field is moving fast.
UPF manufacturers are using tobacco-industry tactics — Milbank Quarterly, February 2026.
Primary source. Gearhardt et al. (Feb 2026), Milbank Quarterly — "From Tobacco to Ultraprocessed Food: How Industry Engineering Fuels the Epidemic of Preventable Disease." University of Michigan / Duke / Harvard.
What the paper says. Five specific tactics the food industry imported from Big Tobacco:
- Dose optimisation — engineering precise sugar / salt / fat ratios for maximum craving
- Delivery speed — fast absorption to spike dopamine rapidly
- Hedonic engineering — the bliss point, engineered sensory reward
- Environmental ubiquity — placement, availability, normalisation
- Deceptive reformulation — "low fat" / "high protein" framing that obscures other harms
Why this matters. Two months old, peer-reviewed, major journal (since 1923), elite institutional authorship. Lets SCANSMART speak like a regulator, not a nutritionist. Direct scaffolding for the structural critique that the manufacturer is the creator of the gap.
Citation. Gearhardt AN et al. Milbank Q. February 2026. DOI via onlinelibrary.wiley.com/doi/10.1111/1468-0009.70066.
UPF intake tied to cognitive decline and stroke.
Primary source. Bhave et al. (May 2024), Neurology (flagship journal of the American Academy of Neurology). Massachusetts General Hospital / REGARDS cohort.
| Finding | Magnitude |
|---|---|
| 10% increase in UPF intake → cognitive impairment risk | HR 1.16 (16% higher) |
| 10% increase in UPF intake → stroke risk | HR 1.08 (8% higher) |
| Stroke risk interaction effect in Black participants | HR 1.15 (higher) |
Why this matters specifically. Plain-English numbers for landing pages, investor decks, NHS pitches. The Black-participant differential aligns with SCANSMART's community-health positioning.
Citation. Bhave VM et al. "Associations Between Ultra-Processed Food Consumption and Adverse Brain Health Outcomes." Neurology 2024. DOI 10.1212/WNL.0000000000209432. PubMed 38776524.
Structural brain changes in 30,000 scans — observational.
Primary source. Kanyamibwa et al. (2025), npj Metabolic Health and Disease. University of Helsinki / Montreal Neurological Institute. UK Biobank, n ≈ 30,000.
What the paper says. High UPF consumption associated with structural changes in three brain regions:
- Hypothalamus — hunger regulation
- Amygdala — emotional response to food
- Right nucleus accumbens — dopamine / reward centre
Authors state these changes "may lead to a cycle of overeating" — feedback loop, not one-way.
Critical caveat (must always be included). Causality NOT confirmed. Cross-sectional, observational. Cannot yet say UPFs cause changes vs people with these brain differences consume more UPFs. If SCANSMART copy ever implies this proves causation, that is overreach and creates regulatory exposure.
Citation. Kanyamibwa C et al. npj Metab Health Dis. 2025. PMC PMC11978510.
Directionally supported, not a single proven chain.
The sequence "engineered foods → blood sugar spikes → inflammation → microglial changes → neuronal changes → faster memory ageing → dementia risk" is a narrative construction joining multiple evidence streams. Each link has evidence; the complete chain is not proven end-to-end.
What is established (and citable)
- UPFs → systemic inflammation: established
- Inflammation → microglial activation → neurodegeneration: established in neurological literature
- UPF consumption → cognitive impairment risk: confirmed (Neurology 2024 above)
- UPFs → structural brain changes: confirmed (Helsinki/UK Biobank 2025 above)
What is NOT established
The full cascade as a single proven mechanism. The evidence stands on each link cited to its own paper — not on a unified causal claim.
Copy-paste-ready primary sources.
- Kanyamibwa et al. npj Metab Health Dis. 2025. PMC PMC11978510.
- Bhave VM et al. "Associations Between Ultra-Processed Food Consumption and Adverse Brain Health Outcomes." Neurology 2024. DOI 10.1212/WNL.0000000000209432.
- Bhave VM et al. PubMed 38776524.
- Mass General Hospital press release (Bhave et al.).
- Gearhardt AN et al. "From Tobacco to Ultraprocessed Food." Milbank Q. Feb 2026.
- Gearhardt AN et al. (Wiley). DOI 10.1111/1468-0009.70066.
- Milbank press release (Gearhardt).
Context reading (not primary sources — for framing, not citation): Michael Moss, Salt Sugar Fat: How the Food Giants Hooked Us (2013); Michael Moss, Hooked: Food, Free Will, and How the Food Giants Exploit Our Addictions (2021).
UK 2026: brain-and-cognition claims sit at the MHRA-medical-device boundary.
Brain-and-cognition dietary claims operate in a particularly tight regulatory space because of the medical-device classification risk. The map below covers what UK regulation permits and where the boundary lies.
| Surface | Mechanism | UK status 2026 | Upstream actor | International parallel |
|---|---|---|---|---|
| Health claims on packs | EFSA Article 13/14 register; UK retained | Tightly regulated. Specific authorised health claims permitted with specific conditions of use. Cognitive-related claims are restricted to a small set of EFSA-authorised statements (e.g., specific iodine or iron claims for cognitive function in children with conditions of use). | Manufacturer making the claim. | EU EFSA Article 13/14; FDA Qualified Health Claims; equivalent jurisdiction-by-jurisdiction. |
| Disease-prevention claims | "Prevents [condition]" / "treats [condition]" framing | Generally prohibited on food. Disease-prevention claims on food products approach Medicines and Healthcare products Regulatory Agency (MHRA) classification territory; foods making medicinal claims may be reclassified as medicines requiring product licensing. | Food manufacturer; advertising agency. | FDA Drug-Food classification; EU Regulation 1924/2006 on nutrition and health claims. |
| Food for Special Medical Purposes (FSMP) | Specialist clinical-indication food products | Tightly regulated. Commission Delegated Regulation (EU) 2016/128 retained; products specifically for dietary management of cognitive or dementia-related conditions require FSMP classification. | Manufacturer of clinical-dietary product. | EU FSMP framework; FDA medical-food regime. |
| SCANSMART editorial register | Two-Layer Literacy Rule first layer; educational not clinical | Educational register. SCANSMART decoder pages, including this brain-and-cognition evidence vault, sit in the educational register: literacy and evidence-base orientation, not clinical-decision-support or medical advice. | SCANSMART editorial team. | Equivalent literacy-versus-clinical-advice distinction across nutrition-literacy platforms internationally. |
| NICE clinical guidelines (cognitive) | NICE NG97 Dementia; equivalent guidance | Clinical guidance. NICE NG97 (Dementia: assessment, management and support for people living with dementia and their carers) provides clinical-pathway dietary advice. Mediterranean and MIND dietary patterns referenced in supporting evidence (decoded in Dietary Patterns). | NICE; clinical specialty bodies. | National dementia / cognitive guidance frameworks globally. |
| HFSS placement and advertising | Indirect cognitive-protection lever via reduced UPF exposure | Restricted (HFSS only). SI 2021/1368 placement (Oct 2022), volume promotions (Oct 2025), ASA/CAP HFSS advertising (Jan 2026). Operates upstream of cognitive-outcome consideration. | Manufacturer; retailer. | See Impulse Buying Triggers and Food Marketing to Kids. |
How to read the map. Cognitive and brain-health claims sit in particularly tight regulatory space because the MHRA medical-device boundary is close. SCANSMART decoders cite the peer-reviewed cognitive-outcome evidence in the educational register without making clinical-decision-support or disease-prevention claims. The HFSS regulatory frame operates upstream of the cognitive-outcome question by reducing UPF exposure at the population level; the cognitive-outcome evidence base supports the HFSS regulatory case rather than the other way around.
Six populations for whom the cognitive-and-UPF evidence base is most relevant.
Three live tensions in the cognitive-and-UPF literature.
1. Causality vs association.
The strongest cognitive-and-UPF evidence (Bhave 2024 Neurology; Kanyamibwa 2025 npj Metab Health Dis) is observational. The associations are real and statistically robust; the causal-direction question is methodologically open. The Kanyamibwa paper explicitly flags that brain-structural differences may pre-date and predict UPF intake rather than result from it (reverse causation). The cascade reading ("UPFs cause cognitive decline through inflammation through microglial activation through structural change") is directionally supported by multiple evidence streams but has not been demonstrated end-to-end as one proven mechanism. The honest reading: the cognitive-outcome evidence supports UPF-reduction as a population-health intervention without proving it as a clinical-treatment intervention.
2. Confounding by dietary pattern.
UPF intake correlates with multiple other dietary-pattern features (low fibre intake; low whole-grain intake; low fruit-and-vegetable intake; low pulse intake; high free-sugars intake; high sodium intake). Statistical adjustment for individual dietary components partially addresses this but cannot fully isolate the UPF-specific contribution from the broader dietary-pattern contribution. The dietary-pattern framing (decoded in Dietary Patterns) is the appropriate level of analysis; the UPF-cognition evidence supports the broader pattern-level conclusion rather than a UPF-specific isolated effect.
3. The "tobacco-tactics" framing as analogy vs identity.
The Gearhardt 2026 Milbank Quarterly paper applies a tobacco-tactics analytical framework to the food industry, identifying five specific tactics (dose optimisation; delivery speed; hedonic engineering; environmental ubiquity; deceptive reformulation). The analogy is academically substantive and the framework is well-supported. The honest reading: the food industry is not identical to the tobacco industry (food is essential to life; tobacco is not), but the documented marketing and engineering tactics show substantial structural similarity. Treating the analogy as identity overclaims; treating it as irrelevant underclaims. The peer-reviewed evidence base supports the analytical framework while acknowledging the contextual differences.
What this brief does not claim.
This evidence vault is written in the educational register and is not clinical-decision-support, personalised dietary advice, medical advice, or a recommendation for the prevention or treatment of any cognitive or neurological condition. Discussion of UPF and cognitive outcomes is general descriptive analysis supported by peer-reviewed sources (Gearhardt et al. 2026 Milbank Quarterly; Bhave et al. 2024 Neurology; Kanyamibwa et al. 2025 npj Metabolic Health and Disease). For any cognitive-related dietary management — diagnosed dementia, neurodegenerative conditions, psychiatric conditions, paediatric developmental considerations — readers should seek input from qualified clinical teams, registered dietitians, and NHS clinical pathways. NICE NG97 (Dementia) and equivalent clinical guidance is the appropriate clinical reference.
This brief contains no allegation of unlawful conduct against any named manufacturer, retailer, or food business operator. The Gearhardt 2026 Milbank Quarterly "tobacco-tactics" framework is presented as the paper's academic analytical claim; the structural-tactics description (dose optimisation, delivery speed, hedonic engineering, environmental ubiquity, deceptive reformulation) is the paper's published academic argument and is cited as such. No factual claim is made about any specific manufacturer's individual conduct beyond what has been peer-reviewed or publicly documented.
MHRA-safety positioning. This brief sits in the educational register, not the medical-device register. SCANSMART is a food literacy and decision-support platform; it is not a medical device and does not provide medical advice. Cognitive-and-UPF dietary-pattern evidence is presented for literacy purposes, not for individual clinical decision-making.
Where to go next.
The full Knowledge Library carries five streams. The peer-reviewed evidence base for the structural critique of industry-funded nutrition research is in Industry Funding Bias in Nutrition Research. The behaviour-change defensibility argument for decision-point capture is in the Behaviour Change & Decision-Point Capture vault. The frozen-aisle expression of the same structural critique is in Frozen Food in the UK. The FSMA gold-standard evidence-vault companions to this brief are Impulse Buying Triggers, Food Marketing to Kids, Brand vs Manufacturer, Reformulation Tracking, Cultural Food Myths (with particular relevance to the Black-participant stroke differential and diaspora-community cognitive risk), Global Staple Foods, Dietary Patterns (including the MIND diet for cognitive health and the Mediterranean-DASH evidence base), Carbohydrate Types, and Caffeine and Health. The NOVA framework that classifies the UPF substrate is in Ultra-Processed Foods.
UPF Brain & Cognitive Claims Evidence Base v1.1 (gold-standard depth) · Compiled 24 April 2026; gold-standard upgrade 11 May 2026 · Stale-date reminder: re-check October 2026 — this field is moving fast · Educational register; not clinical-decision-support; not medical advice.