E120, E904, E920: Animal-Derived Food Additives

Published: 27 September 2026 Updated: 7 October 2026

An E number tells you what a substance does in food and that it passed a safety assessment. It does not tell you what the substance is made from. Four numbers come up again and again when people ask which additives are animal-derived — E120, E904, E920 and E441 — and each answers that question differently. The first is always an insect. The second is an insect secretion. The third can be made three different ways, with nothing on the label to say which. The fourth is not on the additive list at all. We take them in that order.

E120 (carmine, cochineal): always an insect

Carmine is the one case with no ambiguity. The EU specification defines E120 as cochineal extract, carminic acid and carmines, and the dye is extracted from the dried bodies of female cochineal insects (Dactylopius coccus). There is no plant-based version sold under this number. If a label says E120, cochineal, carmine or carminic acid, the colour came from insects.

On safety, EFSA re-evaluated E120 in 2015 and kept the acceptable daily intake set by the earlier committees: 5 mg of carmine per kilogram of body weight per day. Because carmine is roughly 50 per cent carminic acid, the panel restated that limit as 2.5 mg of carminic acid per kg of body weight per day. The panel found no toxicological concern in the animal studies it reviewed and concluded that carminic acid is not genotoxic. It did ask for the EU specification to be renamed and for the limits on toxic element impurities to be tightened.

The clinically important point about carmine is not toxicity but allergy. It is one of the few colours with documented IgE-mediated reactions:

A 1998 case report describes a 35-year-old warehouse worker who developed occupational asthma and rhinoconjunctivitis from handling carmine, and then a reaction after eating a red sweet containing it.
A 2022 study used placebo-controlled oral challenge to assess carmine allergy in urticaria patients with suspected additive hypersensitivity — the reactions are both immediate and delayed, and the authors note they are frequently overlooked.
A 2018 review in Dermatitis describes systemic contact dermatitis to carmine after re-exposure in already sensitised people.
A 2009 report traced an anaphylactic reaction blamed on generic azithromycin to the carmine dye in the tablet coating, not the antibiotic.

That last case is the practical lesson: carmine turns up in tablet coatings, not only in food.

E904 (shellac): the secretion of a scale insect

Shellac is the shiny coating on sweets, chocolate lentils, coffee beans and waxed apples, and on many tablets and capsules. The EU specification is precise about what it is: shellac is purified and bleached lac, the resinous secretion of the insect Laccifer (Tachardia) lacca Kerr.

How it is collected matters to anyone deciding whether to avoid it. The EFSA re-evaluation describes the process: lac insects colonise tree branches, which are cut into roughly 20 cm lengths and harvested as “sticklac”. Sticklac is the branch, the resin, the wax, the polysaccharides — and the insects, which are part of the crushed raw material that is then sieved and refined. So shellac is not simply a secretion gathered from a living animal, the way people sometimes assume by analogy with honey.

EFSA re-evaluated shellac in 2024 and derived an acceptable daily intake of 4 mg per kg of body weight per day for wax-free shellac produced by physical decolouring, from a no-observed-adverse-effect level of 400 mg/kg bw per day with an uncertainty factor of 100. Three details in that conclusion are worth carrying:

The ADI is temporary, pending data the panel asked for on organochlorine impurities.
It does not apply to wax-containing shellac used as a food additive.
In several age groups the ADI was exceeded at the 95th percentile in the higher exposure scenarios. The panel judged the exceedance low and both the exposure and toxicity estimates conservative. Even so, passing the intake limit is an unusual finding for a glazing agent, and reason enough to notice how many glazed products a child eats in a day.

E920: the same number, three different origins

With the first two numbers, the origin is part of the definition. With this one it is not. L-cysteine is a flour treatment agent. It breaks disulfide bonds in gluten, which relaxes dough and shortens kneading — useful in industrial bread, pizza bases, bagels and pastry.

Here the E number genuinely cannot tell you the origin, because the same substance is authorised from several manufacturing routes. EFSA’s 2025 assessment of L-cysteine in animal feed sets the routes out plainly. L-cysteine hydrochloride monohydrate is authorised when produced by fermentation with production strains such as Escherichia coli, and separately when produced by hydrolysis of keratin from avian feathers. L-cysteine itself is authorised when produced by chemical synthesis or protein hydrolysis. Keratin hydrolysis means breaking down feathers, bristles or hair.

Fermentation has become the dominant industrial route, but “dominant” is not “only”, and the label carries the same four characters (E920) either way. A vegan buyer who wants certainty about E920 has two options: a product carrying a vegan certification, or a written answer from the manufacturer about that specific product.

E441: the number that no longer exists

The first three numbers are all on the additive list. The fourth is not. E441 appears on countless “animal ingredients to avoid” lists as gelatine. It is worth being accurate about it: E441 is not on the current list of approved additives. Check the UK Food Standards Agency’s approved additives list, where E120, E901, E904, E920, E471 and E631 all appear with their names: there is no E441.

That is not because gelatine disappeared. It is because gelatine is not regulated as an additive. It is a food ingredient made from animal collagen, governed by the animal-products hygiene rules — the framework EFSA works within when it assesses, for example, the BSE risk of collagen and gelatine from ruminant bones. Its origin is not in doubt: the 2024 EFSA opinion states that pigs, cattle and fish are the principal species used for collagen and gelatine production.

The practical consequence is the opposite of reassuring. Because gelatine is an ingredient rather than an additive, it is named in the ingredients list in plain language — “gelatine”, “jelatin”, “beef gelatine” — and there is no E number to scan for. Searching a label for E441 and finding nothing proves nothing at all.

The grey numbers: E471, E631, E901

Three more numbers deserve a mention: these are the ones wrongly ruled out as animal-derived, and the ones wrongly waved through as plant-based.

E471, mono- and diglycerides of fatty acids. An emulsifier in bread, ice cream, margarine and biscuits. The fatty acids can come from plant oils or from animal fat. The number does not say which.
E631, disodium inosinate. A flavour enhancer, usually paired with E621. It can be produced from meat or fish material, or by fermentation. Again, the number does not say.
E901, beeswax. Always animal-derived in the sense that matters to a vegan — it is made by bees — though many people who avoid meat and dairy treat it differently from insect-derived colours. That is a personal line, not a labelling question.

How to actually read a label

Four habits do most of the work:

Read the ingredients list, not just the E numbers. Gelatine, whey, casein, lactose, carmine and shellac often appear by name. The E-number scan misses all of the ingredients that have no number.
Treat E120 and E904 as settled questions, and E471, E631 and E920 as open ones. The first two are animal-derived by definition. The last three need a certification or a manufacturer’s answer.
Check coatings and glazes, including on medicines. Carmine and shellac are both used on tablets, where the ingredients are listed differently from food.
Use certification where it exists. A vegan seal transfers the verification work to someone who can audit the supply chain, which a shopper reading a label cannot do. Our guide to what the certification labels mean and the comparison of Certified Plant Based and Certified Vegan explain what each seal does and does not promise.

For medicines and supplements specifically, where excipients are rarely obvious and substitution is not always possible, we have a separate guide on spotting animal-derived ingredients in medicines.

What Turkish law requires: the information exists, but not on your label

Turkey’s Food Codex Regulation on Food Additives (Official Gazette, 30 June 2013, no. 28693) contains a provision that bears directly on this article. Listing what must appear on the packaging of the additive itself, two articles require the following:

Article 22, for additives sold business-to-business rather than to the final consumer: “the name of the source from which the food additive is obtained” and “the species of animal from which an animal-origin food additive is obtained”.
Article 23, for additives sold to the final consumer, repeats both: the name of the source, and the animal species for an additive of animal origin.

So whether the E920 in a loaf came from feathers or from fermentation, and whether the fat behind E471 was animal or plant, has to be written down in Turkey — on the additive’s own packaging. Not in the ingredients list of the finished product on the shelf.

The practical consequence: the information you want is recorded somewhere in the chain, and the manufacturer holds it. Writing to a brand to ask what the source of its E471 is, is not asking an unanswerable question; it is asking for something the manufacturer had to receive from its own supplier. Getting no answer does not mean the answer does not exist.

The same regulation also provides for an extra label warning for colours: Article 24 requires additional information on the labels of foods containing certain colours listed in its annex. Whether carmine is on that list is not stated here, because the annex file could not be downloaded; it will be checked separately.

A note on scope

The regulatory statements here refer to EU food additive law — Regulation (EC) No 1333/2008 and the specifications in Commission Regulation (EU) No 231/2012 as cited in the EFSA opinions — and to the UK Food Standards Agency’s published list of approved additives. On the Turkish side only the main text of the Food Additives Regulation was read; the annexes carrying its E-number lists could not be downloaded. No claim is therefore made here about which numbers appear on the Turkish list — the E-number comparisons above rest on the EU and UK lists. We will cover Turkey’s labelling rules separately.

Sources

EFSA Panel on Food Additives and Nutrient Sources added to Food (ANS). Scientific Opinion on the re-evaluation of cochineal, carminic acid, carmines (E 120) as a food additive. EFSA J. 2015;13(11):4288. doi:10.2903/j.efsa.2015.4288.
EFSA Panel on Food Additives and Flavourings (FAF). Re-evaluation of shellac (E 904) as a food additive and a new application on the extension of use of shellac (E 904). EFSA J. 2024;22(8):e8897. doi:10.2903/j.efsa.2024.8897.
EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP). Safety and efficacy of the feed additives l-cysteine and l-cysteine hydrochloride (monohydrate and anhydrous). EFSA J. 2025;23(10):e9689. doi:10.2903/j.efsa.2025.9689.
EFSA Panel on Biological Hazards (BIOHAZ). BSE risk posed by ruminant collagen and gelatine derived from bones. EFSA J. 2024;22(7):e8883. doi:10.2903/j.efsa.2024.8883.
Ministry of Food, Agriculture and Livestock (Turkey). Turkish Food Codex Regulation on Food Additives. Official Gazette. 30 June 2013, no. 28693. Accessed 26 September 2026.
Food Standards Agency. Approved additives and E numbers. Accessed 26 September 2026.
Acero S, Tabar AI, Alvarez MJ, Garcia BE, Olaguibel JM, Moneo I. Occupational asthma and food allergy due to carmine. Allergy. 1998;53(9):897-901. doi:10.1111/j.1398-9995.1998.tb03998.x.
Sadowska B, Sztormowska M, Gawinowska M, Chełmińska M. Carmine allergy in urticaria patients. Postepy Dermatol Alergol. 2022;39(1):94-100. doi:10.5114/ada.2020.100821.
Rundle CW, Jacob SE, Machler BC. Contact Dermatitis to Carmine. Dermatitis. 2018;29(5):244-249. doi:10.1097/DER.0000000000000386.
Greenhawt M, McMorris M, Baldwin J. Carmine hypersensitivity masquerading as azithromycin hypersensitivity. Allergy Asthma Proc. 2009;30(1):95-101. doi:10.2500/aap.2009.30.3199.

Editor: Anıl Güler, Pharmacist
Last scientific review: 27 September 2026