UPF scanner UK — see if a product is ultra-processed in seconds

What "ultra-processed" actually means, and what the scanner shows

Ultra-processed food (UPF) has been the most-Googled UK nutrition story of 2025 and 2026, and the noise around it is genuinely confusing. So before we tell you what our scanner does, here is the plain version of what UPF means, where the term came from, and why we built a tool around it.

UPF is short for the fourth group of the NOVA classification system — a research framework developed by Carlos Monteiro and colleagues at the University of São Paulo. NOVA sorts foods by how much industrial processing they have been through, not by calories or macros. It is widely used in nutrition research and has been picked up by the BMJ, The Lancet, the NHS in its dietary guidance, and a growing number of UK reviews. It is not a UK regulatory standard. The Food Standards Agency does not have a specific UPF policy; it regulates additives one by one. Worth keeping that distinction in mind any time you see a "Group 4" label.

What our app does is read a UK barcode and tell you, in about three seconds, two things: what is in the product (the full ingredients line plus every additive picked out), and where it sits on the processing spectrum (our per-product grade, which factors in additive count, refined ingredients, sweeteners, emulsifiers and the rest). You don't get a NOVA group number stamped on the screen — we don't think single-letter or single-number verdicts are honest enough — but you do get the underlying signals that go into one, in plain English.

The four NOVA groups, with UK examples

Worth being honest about the line between Group 3 and Group 4 — it isn't always crisp. A tinned soup made from vegetables, water and salt is Group 3. The same soup with added emulsifiers, flavourings and modified starch is Group 4. A traditional fresh bread (flour, water, yeast, salt) is Group 3. A long-life sandwich loaf with emulsifiers, preservatives and added enzymes is Group 4. Reasonable people can disagree on borderline cases, and the NOVA group on its own doesn't capture the nutrition picture either. That's why we don't just stamp a group number on the scan — we show the underlying signals.

What you actually see when you scan a UK barcode

What the current research says — and what it doesn't

The most cited piece of UK-relevant evidence in 2024–2026 is the BMJ umbrella review by Lane and colleagues (BMJ, 2024). It pulled together meta-analyses across 32 health outcomes and found higher ultra-processed food intake associated with a wider risk profile — cardiovascular disease, type 2 diabetes, depression, weight gain, and earlier all-cause mortality, among others. That is a serious body of work and it has shifted the UK conversation.

Two things to keep in mind. First, almost all of the underlying studies are observational — they measure what people eat and what later happens to their health, but they cannot prove cause and effect. Diets high in UPF tend to differ from low-UPF diets in many other ways (more calories, more refined sugar, less fibre, less cooking from scratch, often lower household income), and untangling which factor is doing the damage is genuinely hard. Causal pathways — gut microbiome disruption from emulsifiers, blood-sugar effects from refined starch, displacement of fibre and micronutrients — are being studied actively but are not closed cases.

Second, the NHS position remains pragmatic. Current NHS guidance recommends limiting ultra-processed food as part of a balanced diet built on the Eatwell Guide. It does not ban UPF, does not say all UPF is dangerous, and does not endorse any single classification system as the rule. The FSA regulates individual additives — sweeteners, colourings, preservatives — rather than the UPF category as a whole.

What our scanner can NOT do

• It can't diagnose anything. If you have IBS, suspected food intolerances, allergies, a chronic condition, or you're pregnant or managing diabetes, this isn't the place to get medical advice. Track what you eat, take the data to your GP or a registered dietitian, and let them work with you.
• It can't replace reading the back of pack for serious allergies. The 14 UK-recognised allergens (celery, cereals containing gluten, crustaceans, eggs, fish, lupin, milk, molluscs, mustard, peanuts, sesame, soybeans, sulphur dioxide, tree nuts) are exactly what we match against, but on rare unindexed packs the safest move is still to read the label yourself. Where allergens look absent in our data we say "isn't in the ingredients we have", never "this product is free from X".
• It can't tell you a food is "safe" or "fine". Regulatory approval is a regulatory category, not a biological verdict. Our additive copy describes what published research says — "linked to", "may disrupt", "regularly eating this is associated with" — and falls silent when the evidence is unsettled. We don't carry water for manufacturers.
• It can't read loose produce. If there's no barcode on the pack — a sliced ham from a deli counter, a market tomato, a baker's cob — use the AI meal scan on Pro instead. That's a camera-based estimate, not a barcode lookup.
• It can't give you a single number that captures your whole diet. A "20% UPF day" and a "20% UPF day" can look very different depending on what the other 80% is. We surface the per-product picture; the lifestyle interpretation is yours.

Free vs Pro — what you get for £3.99/month

The barcode scan and the per-product grade are free. We don't paywall the thing the page is named after. Free users can log up to 25 foods a day and 5 reaction entries, look up any E-number on the public database at /e-numbers/, and use the additive breakdown without limits.

NutraSafe Pro is £3.99 a month, monthly only (no annual tier yet), iOS only. Here's what it adds:

• Detailed processed-food insights. The per-product grade is free. The full processed-food insights view — additive load over the day, week and month, processing-marker breakdown, week-on-week trends, and how your diet's processing profile is shifting — is the Pro layer.
• Vitamin and mineral tracking against UK NRVs. Aggregate everything you've scanned and logged across the day, week and month, and see whether you're consistently hitting NRVs for vitamin D, folate, iron, B12 and the rest.
• AI Coach. A coach inside the app that's read everything you've scanned, knows your goals and reactions, and can answer questions like "why did I feel rough on Wednesday" with the food log to back it up.
• AI meal scan. No barcode on what you're eating? Take a photo of the plate and we estimate calories, macros and a rough additive profile. Less precise than a barcode scan, more honest than guessing.
• Allergen warning detail. Beyond the simple "this contains" flag, see the exact ingredient sub-string that triggered the match and any may-contain trace warnings the manufacturer has declared.
• Suspected-triggers analysis. Cross-references your reactions against everything you ate in the previous 24–48 hours and surfaces statistically suspicious ingredients, additives or food groups. Designed to give you something concrete to take to your GP — not a diagnosis.
• Fasting tools and full reaction-pattern history.

Who this is for

• Anyone trying to cut back on UPF without going all-or-nothing. The scanner makes the trade-off visible product by product. You don't have to memorise a list of brands.
• Parents reading school-snack packs. The Southampton Six colourings the FSA flagged for hyperactivity in children — sunset yellow (E110), quinoline yellow (E104), carmoisine (E122), allura red (E129), tartrazine (E102), ponceau 4R (E124) — are picked out automatically.
• People tracking food intolerances. Scan, log, cross-reference symptoms later. Pro adds the suspected-triggers analysis on top.
• Households with allergies. Set the household allergens once, every scan flags matches against the ingredients on file. (We say "isn't in the ingredients we have", never "this product is free from X" — liability and accuracy.)
• People who want supermarket honesty. Two yogurts on the same shelf, similar calorie counts, very different ingredient lines and grades. Scan both, see the difference, choose.

How accurate is this, really?

The barcode database is large but not exhaustive. Manufacturers reformulate without re-issuing the barcode, so there is always some lag between the pack you are holding and the data we hold. We display the date the entry was last verified inside the app, and you can submit a correction in two taps if a recipe has moved on — a real human reviews submissions. Where the data we have is thinner than what's on the pack, we say so rather than failing silently. If you are acting on something life-critical (a serious allergy, a medical diet), the back of pack is still the source of truth and we'd never tell you otherwise.

Frequently asked questions

Sources

• Lane MM et al, BMJ 2024 — Ultra-processed food exposure and adverse health outcomes (umbrella review of 32 outcomes).
• NHS — The Eatwell Guide — UK dietary guidance, including limiting ultra-processed food.
• FSA — Food Additives — UK regulator guidance on additives, including the Southampton Six colourings.
• NOVA classification — Monteiro CA et al, University of São Paulo. Research framework, not a UK regulatory category.