E numbers: what that small code on your label really tells you

A number is a name, not a verdict. But that doesn’t make your worries nonsense.

I have built up a kind of trust in the supermarket. Not with the shop, but with the products themselves. That trust didn’t appear out of nowhere: I stood at the shelf with a packet in my hand, read the small print, really read it, and approved or rejected products in my head. One of my filters was simple. Few or no E numbers, then it’s good.

Until I looked closer.

One packet said packaged in a protective atmosphere. What does that even mean? On another I saw E621 and felt my heart rate climb. Neurotoxic, I thought immediately…

What I found was more uncomfortable than my own filter, but also far calmer. On some points I was wrong. On others my distrust was justified, just not in the place I was looking.

First, a currant bun that refused to spoil

I had bought a currant bun. It sat in my bread tin for nearly a month. No mould. Nothing. As if it were immortal.

That unsettled me. What did they do to you, I thought, that you no longer spoil?

The honest answer is that I can never prove anything about that bun, because the label went out with the recycling long ago. But the usual suspect has a name and a number: E282, calcium propionate. It is the classic mould inhibitor in pre-packed bread and baked goods. It slows moulds and the bacterium that turns bread ropy, while leaving the yeast alone, so the dough still rises.

There is also a less exciting possibility: that the bun simply dried out. Mould needs moisture. A dried-out bun isn’t preserved, it’s mummified.

But my unease wasn’t foolish. Keep that name in mind, because calcium propionate returns later in this article, in a place you won’t expect.

An E number is a licence plate, not a seal of approval

Picture an enormous car park with thousands of vehicles. Every car has a licence plate. That plate tells you exactly which vehicle it is, who registered it and when it was last inspected. It tells you nothing about the driver, and nothing about whether that car is dangerous today.

That is how an E number works.

Put simply: it is a registration number. The E stands for Europe. The number means the substance has been assessed, that it appears on the European list of permitted additives, and that rules apply to where and how much it may be used.

An additive is a substance added to food for a technical purpose. Slowing spoilage, keeping fat and water mixed, adding colour, sweetening, preventing caking. That is the whole definition. There is no value judgement in it.

And there sat my first mistake. A label without E numbers does not mean a product without additives. The manufacturer may choose. It may say antioxidant: E300. It may equally say antioxidant: ascorbic acid. Same substance. That’s vitamin C.

So my filter wasn’t filtering for safety. It was filtering for word choice.

Where did this system come from?

Our great-grandparents already practised food chemistry. They just called it something else: drying, salting, smoking, fermenting, pickling.

None of it was harmless. Smoking forms polycyclic aromatic hydrocarbons, some of which are carcinogenic. Plenty of salt slows bacteria and raises your blood pressure. So traditional does not automatically mean safe. That was an uncomfortable but necessary correction for me.

The E number system itself was born in 1962, inside what was then the European Economic Community, and it started with colours. Preservatives followed in 1964, antioxidants in 1970, emulsifiers and thickeners in 1974.

The goal was not to hide anything. It was the opposite. Six countries with six languages and six names for the same substance could not check each other’s products. A number is language-free. E330 is E330 in Lisbon, in Athens and in Amsterdam.

The inspection, and why cars sometimes get taken off the road

In Europe, EFSA does the scientific assessment; the European Commission and the member states take the decision.

The key figure is the ADI: the acceptable daily intake. In plain language: the amount you could take in every single day, your entire life, without a meaningful health risk being expected.

How is that figure reached? Researchers find the highest dose at which no harmful effect was seen. Then they divide it by one hundred. Ten times for the difference between lab animal and human, ten times for the difference between one human and another.

Think of the ADI as a speed limit, not the speed at which you crash. Doing 105 km/h once is not a disaster. Doing 140 km/h every day is another matter.

And then the inspection. An E number is not a permanent approval. Additives must be reassessed with today’s knowledge, and sometimes they don’t survive it.

Example: E171, titanium dioxide. For years it was the white colour in chewing gum, sweets and icing. In 2021 EFSA said genotoxicity could not be ruled out.

In plain language. Your DNA is the manual that tells every cell how to behave. Genotoxic means: this substance can damage letters in that manual. And with that kind of damage you often cannot point to a threshold below which it definitely doesn’t happen. No threshold means no safe daily amount can be written down. And there it stops.

In January 2022 the EU withdrew the authorisation. The car came off the road.

Oxygen has a licence plate

Back to that packet in my hand.

Packaged in a protective atmosphere means something far duller than I imagined. Air has been removed from the pack and replaced by a gas mixture that slows spoilage. Those gases all have E numbers: E941 nitrogen, roughly 78 per cent of the air you are breathing right now. E948 oxygen, yes, plain oxygen; in meat packs it helps preserve the red colour. E290 carbon dioxide, the bubbles in your sparkling water.

So there I stood in the supermarket, having doubts about oxygen. I recommend that moment to everyone.

There are more numbers that sound more alarming than the substance they name:

Number	What it is	Where you know it from
E100	Curcumin	The yellow of turmeric
E101	Riboflavin	Vitamin B2
E160a	Beta-carotene	The orange in carrots
E260	Acetic acid	Vinegar
E270	Lactic acid	Yoghurt, fermentation
E300	Ascorbic acid	Vitamin C
E322	Lecithins	Egg yolk, soy
E330	Citric acid	Citrus, and your own citric acid cycle
E406	Agar	Gelling agent from seaweed
E440	Pectin	The fibre in apples that sets jam

Citric acid is industrially produced by fermentation rather than squeezed from a lemon. That doesn’t change the molecule. Your body can’t tell the difference, because there isn’t one.

And the reverse lesson matters just as much: natural is no proof of safety. Botulinum toxin is natural. Arsenic is natural. Many mould toxins are natural. Safety is decided by the substance, the dose, the duration and the person. Never by origin alone.

What is a manufacturer allowed to leave off?

This is the question you only start asking after reading labels for a while. Is everything really on there?

No. And it isn’t sneaky, it’s written into the law.

Carry-over. If an additive is present only because it was already in one of the ingredients, and it no longer has a technical function in the finished product, it doesn’t have to appear on the label. The logic: a preservative in a sauce that ends up inside a filled product no longer does anything in that final product.

Processing aids. Substances used during production that have finished their job don’t have to be listed either. Think of something that helps with filtering or releasing a product from a mould.

Collective names. The word flavouring can stand on its own. Dozens of substances may sit behind that single word. That is legal, and it is exactly why the length of an ingredient list is a poor guide.

But there is one important exception, and it exists for good reason: allergens must always be declared, even when they arrive through carry-over or as a processing aid. Peanut, gluten, and sulphites above a certain threshold. That is where the exemption ends.

So: what you read on a label is not the complete chemical inventory. It is the list of what has a technical function in the finished product, plus everything that could be dangerous for a sensitive group. That is less than you hope, and more than nothing.

Where the distrust does hold up

There is no honest list of the five most toxic E numbers. There are, however, groups where the concerns are concrete.

Nitrite and nitrate, E249 to E252

Mainly in ham, bacon, salami and sausage. They slow dangerous bacteria, including the one that produces botulinum toxin. That is a real, life-saving function.

But nitrite can contribute to the formation of nitrosamines, and some of those damage DNA. In 2023 EFSA concluded that exposure to nitrosamines is a health concern across all age groups. Separately, IARC, the WHO’s cancer research agency, has classified processed meat as carcinogenic to humans: every 50 grams a day is associated with roughly an 18 per cent higher risk of bowel cancer.

That percentage frightens people, so let’s convert it. Eighteen per cent is a relative risk. Cancer Research UK put it in absolute terms: of every 1,000 people, around 61 will develop bowel cancer at some point, and among those eating the least processed meat that figure is around 56.

What this means: the slice of ham at a birthday is not the issue. The daily habit is.

Sulphites, E220 to E228

Wine, dried fruit, fruit juice, potato products, beer.

In 2022 EFSA reached an uncomfortable conclusion: there was too little data to establish a reliable ADI, and high consumers remain a concern. Sulphites can trigger breathing problems in some people with asthma. That is why, above 10 mg per kilo or litre, they must be declared.

What this means: if you have asthma and recognise reactions after wine or dried apricots, this is not theory for you.

Phosphates, E338 to E341 and E450 to E452

In processed cheese, cured meats, soft drinks, ready meals.

In 2019 EFSA set a group ADI of 40 mg phosphorus per kilo of body weight per day, and calculated that infants, toddlers and children can already exceed it at average consumption. A detail that often gets lost: that ADI does not apply to people with moderate to severe kidney impairment.

What this means: healthy and adult? Not a daily worry. Kidney disease? Then phosphate from additives weighs more heavily than phosphate from ordinary food, because the body absorbs it more easily.

The six colours carrying a mandatory warning

E102, E104, E110, E122, E124 and E129. Products containing these colours must state in the EU that they may have an adverse effect on activity and attention in children.

That warning comes from a single study: the Southampton trial, published in The Lancet in 2007. Almost three hundred children were given a drink for six weeks containing a mixture of colours and sodium benzoate, or not. The drinks looked identical. A small effect on behaviour and attention was found, not the same in every child.

What this means: it is a precaution, not proof that these colours cause ADHD. And brightly coloured sweets have little to offer anyway.

Glycerol in slush drinks, E422

This is the most instructive example of them all.

Glycerol has low toxicity. But in sugar-free slush it is concentrated, and small children drink it fast and cold. After hospital admissions, the UK Food Standards Agency advised in 2025: none at all under the age of seven, and no more than one 350 ml cup a day for children aged seven to ten.

What this means: a substance can be properly approved and still go wrong in one specific product form, at one specific age, in one specific amount. The licence plate was in order. The driving was not.

E621: what actually holds up

This is the number that raised my heart rate. So let me be precise.

What it is. E621 is monosodium glutamate, MSG. Glutamate is an amino acid that occurs naturally in tomato, Parmesan, soy sauce and in your own body. In water, MSG separates into a sodium particle and a glutamate particle. From that moment your body cannot tell whether the glutamate came from a sachet or a tomato.

Why people are afraid. The reasoning sounds logical: glutamate is an excitatory signalling substance in your brain, too much glutamate damages nerve cells, therefore eating MSG damages your brain.

That reasoning skips three doors. Your gut cells burn a large share of dietary glutamate immediately as fuel. Gut wall and liver process more of it before it reaches your general circulation. And the blood-brain barrier strictly controls how much gets into your brain at all.

Nerve cell damage from excess glutamate is real. It matters in oxygen shortage in the brain. But that happens at pathologically high concentrations inside the nervous system, not from half a gram of powder over your soup.

Why animal research looks so alarming. Many classic MSG studies injected the substance in extremely high doses under the skin or into the abdominal cavity of newborn rodents. An injection bypasses gut and liver entirely. That research says something about injections. Not about your soup.

What has been reported. The FDA summarises it this way: sensitive people may experience mild, short-lived symptoms after around 3 grams of MSG or more, taken without food. A normal portion of food containing added MSG usually contains less than 0.5 grams.

And now the part reassuring articles leave out. EFSA reassessed the glutamates in 2017 and set a group ADI of 30 mg per kilo of body weight per day. In doing so, EFSA noted two things. Average exposure exceeded that ADI in infants, toddlers, children and adolescents. And in heavily exposed children, intake approached doses that research has associated with headache.

So: proven carcinogenic? No. Proven brain poison at normal dietary levels? No. Harmless, then? Also no. EFSA saw enough reason for a cautious limit, and saw children going over it.

E621 is not inert. What counts is how often, how much, and at what age.

One more thing, and it is the finest reversal in this whole article. MSG is around 12 per cent sodium; table salt around 39 per cent. In a reformulated recipe MSG can therefore help reduce salt. But in instant noodles it usually sits beside a mountain of salt. Same substance, two opposite roles. Which is exactly why a number tells you so little.

Smoke flavourings: right instinct, wrong address

I thought smoke flavouring, liquid smoke, was an E number. It isn’t. Smoke flavourings fall under the flavourings rules, not the additives rules, and therefore carry no E number.

And that is precisely where the story is. In 2023 EFSA assessed the smoke flavourings on the European market and could not dispel concerns about genotoxicity. In 2024 the EU decided not to renew authorisation for eight of them. They are being phased out: two years where they are added purely for flavour, such as in soups, crisps and sauces, and five years where they genuinely replace smoking, such as in ham, fish and cheese.

Read that again. The substance the regulator was most worried about was never on the label as an E number at all.

Europe, America, and your holiday suitcase

The EU works with a positive list: a substance may only be used if it is explicitly on that list, in the stated categories and amounts.

The United States additionally has the GRAS route, generally recognized as safe. And you might reasonably ask: why the hurry, is that marketing?

The beginning was actually very practical. In 1958 America introduced a law requiring new additives to be approved before use. The regulator immediately risked drowning in applications for salt, vinegar and pepper: substances humanity had eaten for centuries. GRAS was the exception for that kind of obviousness.

The problem came later. Since 1997, notifying a GRAS decision has been voluntary. A company may therefore conclude, using its own experts, that a substance is safe, and need not tell the regulator. As a result there are probably more than a thousand substances in the American food chain whose assessment only the manufacturer has seen. That is not marketing. It is a system designed to prevent delay, which turned out to also provide a door.

What does that mean on holiday? Not that American food is unsafe. But the systems differ, and not always in the same direction:

• Potassium bromate and azodicarbonamide in flour and bread: not permitted in the EU, permitted within limits in the US.
• Brominated vegetable oil in soft drinks: never permitted in the EU, withdrawn in the US only in July 2024.
• Red dye No. 3: withdrawn for food in the US in January 2025, with a compliance deadline of January 2027.
• Titanium dioxide, E171: banned by Europe, still in use elsewhere.

Practically: a product from another continent isn’t automatically unsafe, but the same brand may be formulated differently there. The ingredient list remains your best source. Outside Europe you simply read names and colour numbers instead of E numbers.

What the newest research shows

Something has shifted in this field: researchers no longer look at single substances, but at mixtures, the way people actually eat them.

The French NutriNet-Santé cohort followed more than 100,000 adults. In 2025 it found a link between two common additive mixtures and more type 2 diabetes. In January 2026 it reported thirteen preservatives associated with type 2 diabetes. And that list included not only sodium nitrite, but also citric acid, sodium ascorbate, alpha-tocopherol, and yes, calcium propionate. The currant bun substance.

Does that mean vitamin C causes diabetes? Almost certainly not, and the researchers don’t claim it does.

Look at their own figures. On average, 1 per cent of tocopherol intake came from additives, 17 per cent of citric acid, and 63 per cent of sulphites. For many of these substances the additive is a highlighter, not a culprit: it marks how heavily processed someone’s diet is.

Add to that the fact that this is observational research. It shows association, not cause, and it comes from a single French cohort.

The honest summary fits in two sentences. There are enough signals to justify better research and reassessment. There is not enough evidence to say these substances cause disease on their own.

Six rules for the shelf

1. Look at the product first, the numbers second. A wholegrain loaf with one emulsifier beats an E-number-free biscuit full of sugar and butter. Every time.

2. Watch repetition, not incidents. The exposure that matters rarely comes from one packet. It comes from what goes into the trolley every week.

3. Limit processed meat as a habit. Better evidenced than any list of E numbers. The charcuterie board on Sunday can stay.

4. Take the specific warnings seriously. Sulphites with asthma. Phenylalanine with phenylketonuria. Sorbitol and maltitol with gut complaints; they act as a laxative, and you will notice. The six colours with sensitive children. Phosphates with kidney disease. Slush under seven.

5. Compare within the same category. Two yoghurts, equal nutrition, one with fewer superfluous extras: fine choice. Not because it is proven safer, but because superfluous is, after all, superfluous.

6. Ask a better question. Not is E330 harmful?, but: why is this in here, how often do I eat this sort of thing, and what is it displacing from my diet?

That last question changes a shopping trolley faster than memorising a hundred numbers.

This article provides general information and is not medical, dietary or psychological advice. If you have questions about your own health, nutrition or wellbeing, please speak with your GP, a dietitian or a treating specialist.

What this digging gave back

Calm, mostly. The number is a registration code, not a verdict. The inspection continues, and something really was taken off the road when the knowledge changed. But it isn’t blind trust either, because plenty is still wrong.

Some of the older additives are still waiting to be reassessed against modern standards. An ADI is set per substance, while we eat mixtures. Children exceed the limit for some substances. And the substance the regulator worried about most wasn’t even on the label as an E number.

But I got back something more valuable than a list of forbidden numbers: I now know where to look. From today, so do you, using exactly the same labels you have always held in your hands.

Preventing illness is easier than curing it. But prevention doesn’t begin with fear. It begins with understanding.

Which habit in your shopping trolley repeats most often, and do you know why it’s there?

---

Related reading

• Chocolate: from sacred drink to ultra-processed snack
• The Living Kitchen: what fermented foods really do to your body
• Honey: superfood or just sugar?
• Sugar: the reward that never has to wait

---

Sources

1. Regulation (EC) No 1333/2008 on food additives. (2008). EUR-Lex. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32008R1333
2. European Commission. (n.d.). Food additives: EU rules and re-evaluation programme. https://food.ec.europa.eu/food-safety/food-improvement-agents/additives/re-evaluation_en
3. Regulation (EU) No 257/2010 setting up a programme for the re-evaluation of approved food additives. (2010). EUR-Lex. https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:02010R0257-20210327
4. EFSA Panel on Food Additives and Flavourings (FAF). (2021). Safety assessment of titanium dioxide (E171) as a food additive. EFSA Journal, 19(5), Article e06585. https://doi.org/10.2903/j.efsa.2021.6585
5. Regulation (EU) 2022/63 as regards the food additive titanium dioxide (E 171). (2022). EUR-Lex. https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32022R0063
6. EFSA Panel on Food Additives and Flavourings (FAF). (2025). Re-evaluation of oxygen (E 948) and hydrogen (E 949) as food additives. EFSA Journal, 23(8), Article e9595. https://doi.org/10.2903/j.efsa.2025.9595
7. Regulation (EU) No 1169/2011 on the provision of food information to consumers. (2011). EUR-Lex. https://eur-lex.europa.eu/eli/reg/2011/1169/oj
8. EFSA Panel on Contaminants in the Food Chain (CONTAM). (2023). Risk assessment of N-nitrosamines in food. EFSA Journal, 21(3), Article e07884. https://doi.org/10.2903/j.efsa.2023.7884
9. International Agency for Research on Cancer. (n.d.). Carcinogenicity of the consumption of red meat and processed meat: Questions and answers. World Health Organization. https://www.who.int/news-room/questions-and-answers/item/cancer-carcinogenicity-of-the-consumption-of-red-meat-and-processed-meat
10. Cancer Research UK. (n.d.). Processed and red meat and cancer. https://www.cancerresearchuk.org/about-cancer/causes-of-cancer/diet-and-cancer/does-processed-and-red-meat-cause-cancer
11. EFSA Panel on Food Additives and Flavourings (FAF). (2022). Follow-up of the re-evaluation of sulfur dioxide (E 220), sodium sulfite (E 221), sodium bisulfite (E 222), sodium metabisulfite (E 223), potassium metabisulfite (E 224), calcium sulfite (E 226), calcium bisulfite (E 227) and potassium bisulfite (E 228). EFSA Journal, 20(11), Article e07594. https://doi.org/10.2903/j.efsa.2022.7594
12. EFSA Panel on Food Additives and Flavourings (FAF). (2019). Re-evaluation of phosphoric acid–phosphates, di-, tri- and polyphosphates (E 338–341, E 343, E 450–452) as food additives and the safety of proposed extension of use. EFSA Journal, 17(6), Article e05674. https://doi.org/10.2903/j.efsa.2019.5674
13. McCann, D., Barrett, A., Cooper, A., Crumpler, D., Dalen, L., Grimshaw, K., Kitchin, E., Lok, K., Porteous, L., Prince, E., Sonuga-Barke, E., Warner, J. O., & Stevenson, J. (2007). Food additives and hyperactive behaviour in 3-year-old and 8/9-year-old children in the community: A randomised, double-blinded, placebo-controlled trial. The Lancet, 370(9598), 1560–1567. https://doi.org/10.1016/S0140-6736(07)61306-3
14. Food Standards Agency. (2025). Glycerol in slush ice drinks: Advice for children. https://www.food.gov.uk/safety-hygiene/glycerol
15. EFSA Panel on Food Additives and Flavourings (FAF). (2026). Safety of acute exposure to the food additive glycerol (E 422) from beverages. EFSA Journal, 24(5), Article e10057. https://doi.org/10.2903/j.efsa.2026.10057
16. U.S. Food and Drug Administration. (n.d.). Questions and answers on monosodium glutamate (MSG). https://www.fda.gov/food/food-additives-petitions/questions-and-answers-monosodium-glutamate-msg
17. EFSA Panel on Food Additives and Nutrient Sources added to Food (ANS). (2017). Re-evaluation of glutamic acid (E 620), sodium glutamate (E 621), potassium glutamate (E 622), calcium glutamate (E 623), ammonium glutamate (E 624) and magnesium glutamate (E 625) as food additives. EFSA Journal, 15(7), Article e04910. https://doi.org/10.2903/j.efsa.2017.4910
18. European Food Safety Authority. (n.d.). Smoke flavourings. https://www.efsa.europa.eu/en/topics/topic/smoke-flavourings
19. Commission Implementing Regulation (EU) 2024/2067 removing smoke flavouring primary products SF-001 to SF-010 from the EU list. (2024). EUR-Lex. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32024R2067
20. U.S. Food and Drug Administration. (n.d.). FDA’s approach to the GRAS provision: A history of processes. https://www.fda.gov/food/generally-recognized-safe-gras/fdas-approach-gras-provision-history-processes
21. U.S. Food and Drug Administration. (2024). Revocation of authorization for use of brominated vegetable oil in food. Federal Register. https://www.federalregister.gov/documents/2024/07/03/2024-14300/revocation-of-authorization-for-use-of-brominated-vegetable-oil-in-food
22. U.S. Food and Drug Administration. (2025). FDA to revoke authorization for the use of Red No. 3 in food and ingested drugs. https://www.fda.gov/food/hfp-constituent-updates/fda-revoke-authorization-use-red-no-3-food-and-ingested-drugs
23. Payen de la Garanderie, M., Hasenböhler, A., Dechamp, N., Javaux, G., Szabo de Edelenyi, F., Agaësse, C., De Sa, A., Bourhis, L., Porcher, R., Pierre, F., Coumoul, X., Kesse-Guyot, E., Allès, B., Fezeu, L. K., Cosson, E., Tatulashvili, S., Huybrechts, I., Hercberg, S., Deschasaux-Tanguy, M., … Touvier, M. (2025). Food additive mixtures and type 2 diabetes incidence: Results from the NutriNet-Santé prospective cohort. PLOS Medicine, 22(4), Article e1004570. https://doi.org/10.1371/journal.pmed.1004570
24. Hasenböhler, A., Javaux, G., Payen de la Garanderie, M., Szabo de Edelenyi, F., Bourhis, L., Agaësse, C., De Sa, A., Huybrechts, I., Pierre, F., Coumoul, X., Julia, C., Kesse-Guyot, E., Allès, B., Fezeu, L. K., Hercberg, S., Deschasaux-Tanguy, M., Cosson, E., Tatulashvili, S., Chassaing, B., … Touvier, M. (2025). Associations between preservative food additives and type 2 diabetes incidence in the NutriNet-Santé prospective cohort. Nature Communications, 16, Article 11199. https://doi.org/10.1038/s41467-025-67360-w
25. Ragesoss. (n.d.). Monosodium glutamate crystals [Photograph]. Wikimedia Commons. Used under CC BY-SA 3.0. https://commons.wikimedia.org/wiki/File:Monosodium_glutamate_crystals.jpg

Sources on this site are cited in APA style, with a DOI wherever one exists. Why I made that choice →