The food additives worth knowing about. And the hundreds you can ignore.
You read the back of a pack, see a row of E-numbers, and assume the worst. Most of them are mundane: vitamin C, citric acid, fruit fibre. A smaller set carry genuine, evidenced concerns. Here is which is which, and what the regulators actually say.
What an E-number actually means
An E-number is not a danger rating. It is an approval code: the additive has been assessed and approved for use in food, and the E just records that it passed that gate. In the EU the assessor is the European Food Safety Authority (EFSA); in Great Britain it is the Food Standards Agency (FSA).
The number itself is a filing system by function, not by risk. The 100s are colours. The 200s are preservatives, the things that hold back mould and bacteria. The 300s are antioxidants and acidity regulators, which stop fats turning rancid and keep a product's acidity steady. The 400s are thickeners, gelling agents and emulsifiers, the texture group. The 500s are more acidity regulators and anti-caking agents. The 600s are flavour enhancers. The 900s are a mixed drawer of glazing agents, packaging gases and most of the sweeteners.
The honest part most labels leave unsaid: approved is not the same as no concern. Approval reflects the evidence at the time it was granted, and the science moves. A handful of long-approved additives are now under active review or have lost their authorisation in one market while keeping it in another. The E tells you something cleared the gate. It does not tell you the gate has not since moved.
Why E-numbers got a bad name
The fear of E-numbers is mostly a story about a few colours, told until it covered all 500 codes. There is a real kernel inside it, and a lot of overspill.
The turning point was a 2007 study from the University of Southampton, funded by the FSA and published in The Lancet. It gave children mixtures of six artificial colours plus the preservative sodium benzoate (E211), and reported a small rise in hyperactive behaviour in some of them. The headlines did the rest. By 2008 the FSA had asked manufacturers to phase the six colours out voluntarily, and many reformulated to natural colours rather than carry a warning line. From 2010 an EU-wide rule made that warning mandatory on any food still using them.
That is the kernel of truth: a specific, regulator-backed finding about a specific set of colours. The overspill is everything that got tarred with it. Vitamin C does not become suspicious because it has a code. Citric acid is not a chemical secret. The useful move is not to fear the letter E, but to learn the short list that actually earns attention and let the rest be what they are: ordinary ingredients with a filing number.
The ones you can mostly ignore
These turn up on thousands of labels. They are ordinary ingredients that happen to have a code, and knowing what they are saves you from worrying about the wrong thing.
Common, but worth knowing what they are
Not flagged for harm, but with a detail behind the code that matters if you have an allergy, follow a particular diet, or just want to know what you are eating.
The ones with real evidence behind the concern
A short list, each backed by a regulator's position rather than a rumour. This is where reading the label earns its keep.
The colours and children
This is the one concern that put a warning on UK shelves. It is worth understanding what the study actually found, and what it did not.
The 2007 Southampton trial (McCann et al., The Lancet) tested two mixtures of additives on children aged three and eight to nine: six artificial colours and the preservative sodium benzoate. The six colours, now called the Southampton Six, are tartrazine (E102), quinoline yellow (E104), sunset yellow (E110), carmoisine (E122), ponceau 4R (E124) and allura red (E129).
The finding was modest. EFSA, reviewing the same data, concluded the study offered limited evidence of a small effect on the activity and attention of some children, and noted the results were not consistent across the two age groups or the two mixtures. It was not strong enough for EFSA to change the colours' acceptable intakes. It was enough for regulators to act on the precaution: in Great Britain the six were phased out voluntarily, and across the EU any food still containing them must carry the line "may have an adverse effect on activity and attention in children". The warning is on the label so a parent can choose. It is not a finding that the colours cause hyperactivity in every child.
Nitrites and processed meat
This is the firmest concern on the list, and the one most muddled by clever marketing. The headline is not the additive code, it is the meat.
In 2015 the WHO's cancer agency (IARC, Monograph 114) classified processed meat as Group 1, carcinogenic to humans, on sufficient evidence for bowel cancer. Group 1 is about the strength of the evidence, not the size of the risk: it sits processed meat in the same evidence tier as tobacco without claiming the same danger. The mechanism points partly at the curing salts, sodium nitrite (E250) and nitrate (E251). Under heat and stomach acid, nitrite can react with proteins in the meat to form nitrosamines, compounds linked to bowel cancer. The NHS advice that follows is to cap processed meat at 70g a day.
The marketing wrinkle is "nitrite-free" or "naturally cured" bacon. In most cases the cure has simply switched to celery powder, which is naturally high in nitrate that the process turns into nitrite. The body cannot distinguish nitrite from celery from nitrite from a packet, and the same nitrosamine chemistry applies. The FSA's position is that there is no evidence celery-cured processed meat is any safer. The honest reading: it is the processed meat that carries the Group 1 label, and a different spelling on the cure does not change that.
The sweetener question
The low-calorie sweeteners attract more rumour than almost any additive group. Here is where the named bodies actually stand on each one.
Aspartame (E951). In July 2023 IARC classified aspartame as Group 2B, possibly carcinogenic, on limited evidence. In the same week the WHO's JECFA committee reviewed the intake side and left the acceptable daily intake unchanged at 40mg per kg of body weight. By WHO's own worked example, a 70kg adult would need roughly 9 to 14 cans of a typical diet drink a day, with no aspartame from any other source, to pass that limit. People with the inherited condition PKU must avoid aspartame because they cannot process one of its breakdown products.
Sucralose (E955). A 2023 laboratory study reported that a compound called sucralose-6-acetate, which can form when sucralose is digested and can appear as a trace impurity, showed signs of genotoxicity in human cells in a dish. Separate work has questioned what happens when sucralose is heated to baking temperatures. These are early in-vitro signals, not a regulator's verdict, and EFSA has not changed its position on sucralose on the back of them.
Acesulfame K (E950) and stevia (E960). Acesulfame K is a synthetic sweetener often blended with aspartame or sucralose; it carries an EFSA acceptable intake and no current alarm. Stevia is extracted from a plant, which some packs lean on, though as the section below notes, plant origin tells you little about effect.
The point that cuts across all of them came in May 2023, when the WHO advised against using non-sugar sweeteners for weight control. Its review found no long-term benefit for body fat in adults or children, and flagged possible links to type 2 diabetes and heart disease with long-term use. The recommendation does not apply to people with existing diabetes. WHO's suggested route to less sugar is unsweetened food and the natural sugars in fruit, rather than swapping one additive for another.
Emulsifiers and the gut
This is the live research frontier, and the one most prone to being oversold. The honest summary: suggestive, not settled, and strongest for sensitive guts.
Emulsifiers keep oil and water from separating, which is why they are in ice cream, low-fat spreads, sauces and a lot of bread. The ones in the research are mono- and diglycerides (E471), carboxymethyl cellulose or cellulose gum (E466) and polysorbate 80 (E433). In 2015 Chassaing and Gewirtz reported in Nature that two of these emulsifiers, fed to mice, thinned the protective mucus layer of the gut, shifted the gut bacteria and triggered low-grade inflammation, with weight and metabolic changes following. Mice are not people, and the doses are debated.
On the human side, the French NutriNet-Sante cohort, which follows tens of thousands of adults, has reported associations between higher intake of certain emulsifiers and a raised risk of some cancers, cardiovascular disease and type 2 diabetes. A cohort shows a link, not a cause: the people eating the most emulsifiers also tend to eat the most ultra-processed food, and untangling the two is hard. Taken together it is a genuine signal worth watching, likely to matter most to people with inflammatory bowel conditions or a sensitive gut, and not yet a finding any regulator has acted on.
Natural does not mean safer
The natural-versus-synthetic split is the label cue people trust most, and it tells you the least about whether an additive matters.
Cochineal (E120), the red in some yoghurts and sweets, is about as natural as a colour gets: it is made from crushed insects. It is also a recognised, if rare, cause of allergic reactions, and it is neither vegan nor vegetarian. Annatto (E160b), another plant-derived colour, has its own occasional sensitivity reports. Plant origin did not spare either of them.
It runs the other way too. The vitamin C added as E300 is ascorbic acid whether it is fermented in a vat or squeezed from an orange: the molecule is identical and the body cannot tell them apart. "Made in a factory" is not a property of the chemistry. The thing that decides whether an additive is worth your attention is the evidence about that specific substance, from EFSA, the FSA, IARC and the rest, not the story on the front of the pack about where it came from.
How to read the label
Once you know the short list, the back of a pack gets quick to scan. Additives sit in a predictable place and follow a fixed grammar.
By law each additive is listed by its function, then its name or its number: "colour: tartrazine", "preservative: E211", "emulsifier: E471". The function word tells you what drawer it came from, and you can use either the name or the code to look it up. Additives appear within the ingredients list, in descending order of weight like everything else, so an additive near the end is present in a small amount.
What to actually scan for, rather than the whole row: the mandatory colour warning line if you are buying for children; the curing salts on processed meat, alongside the 70g NHS guide; the allergen declarations, which the law already puts in bold (soya in lecithin, sulphites in dried fruit and wine). The rest of the codes are mostly the ordinary group, there to do a job the recipe needs. Reading the label well is knowing which handful to slow down for, not reading every E aloud.
Look any code up yourself
A code on its own tells you nothing about whether it matters. The point is to know which handful are worth your attention, then check the rest only when you are curious. You can search all 459 additives and read what each one is, where it comes from and what the regulators say, or start with how the E-number system works.
How NutraSafe helps
This is the gap the app is built for: you have read the back of a pack and you are still none the wiser.
Point your phone at the barcode and NutraSafe reads the ingredients, then pulls every additive it recognises into plain English: what it is, what it is doing in the food, and what the regulators actually say, with the source underneath. The handful worth slowing down for are flagged, so you are not standing in the aisle googling a code to work out whether it matters.
It works on exactly the things this guide is about. The warning colours, the curing salts in processed meat, the sweeteners, the emulsifiers, each comes back with the regulator's position rather than a rumour from a forum, and a tap opens the full evidence and the sources behind it. The additive flag is free on every food it recognises. Pro adds the AI meal scan for a dish that has no barcode, the full ingredient breakdown, and the AI Coach.
Know what is in it, in three seconds
Scan a pack and NutraSafe gives you every additive in plain English, with the source.
Get NutraSafe on the App StoreSources
- EFSA, re-evaluations of food colours, sweeteners and additives, and the EFSA assessment of the McCann et al. (2007) study (EFSA Journal).
- McCann et al., The Lancet 2007, the University of Southampton study on colours and children's behaviour.
- Regulation (EC) No 1333/2008 on food additives, the mandatory warning-label provision for the six colours; Food Standards Agency guidance on the Southampton Six.
- WHO IARC Monographs Volume 114 (2015), red and processed meat (Group 1); Volume 101 (2013), 4-methylimidazole (Group 2B).
- NHS, advice on red and processed meat (70g a day); FSA guidance on nitrites and celery-cured meat.
- EFSA, scientific opinion on titanium dioxide (E 171) as a food additive (2021); EU withdrawal of authorisation, August 2022; FSA interim position for Great Britain.
- IARC and WHO/JECFA, aspartame hazard and risk assessments, July 2023 (Group 2B; ADI 40mg/kg unchanged).
- WHO, guideline on the use of non-sugar sweeteners, May 2023.
- Chassaing et al., Nature 2015, dietary emulsifiers and the mouse gut; NutriNet-Sante cohort analyses (PLOS Medicine 2024; Lancet Diabetes & Endocrinology 2024).
- EFSA, re-evaluation of carrageenan (E 407) as a food additive (2018), temporary acceptable daily intake.