E numbers on Russian shelves: an analysis of 48,000 ingredient lists
We analysed 48,230 ingredient lists from the product pages of six Russian retail chains. A code of the form E-xxx turned up in 22.8 % of them — and that turned out to be not the share of products containing additives, but the share where the additive was written as a code. The additive itself is present in at least 61.4 %.
How the analysis was done
The source is our mirror of product cards from six Russian retail chains, kept so the app can find a product by its barcode. The sample includes national and regional chains, hypermarkets and convenience stores. The mirror holds 124,463 cards, of which 48,230 contain a parsable ingredients string — those are the ones we used.
In each list we looked for two things: a code of the form E-xxx, and the name of an additive spelled out. Codes were matched with both the Latin E and the Cyrillic Е, with a space or hyphen between the letter and the digits. Names were matched against a list of about sixty wordings — "citric acid", "lecithin", "carrageenan", "potassium sorbate" and so on.
Four limits, without which the figures below cannot be read.
- These are retailer product pages, not the whole shelf. The sample contains what a chain published online and filled in an ingredients field for. Loose and fresh goods are barely represented.
- The ingredients were transcribed by the retailer from the pack. Transcription errors are possible and invisible to us.
- The list of names is deliberately incomplete. It excludes flavourings: the regulation has them written as the word "flavouring" without a code, and counting them alongside the rest would be a stretch. The share of lists containing additives is therefore a lower bound, not an exact value.
- We do not compare the chains with each other. Their assortments differ, and so does the share that writes codes — more on that below. Any ranking built on this data would rank labelling style, not composition.
First finding: a code and an additive are different things
| What we looked for | Lists | Share |
|---|---|---|
| Contains a code of the form E-xxx | 11,003 | 22.8 % |
| Contains an additive written as a name | 27,484 | 57.0 % |
| Contains either of the two | 29,617 | 61.4 % |
| Has an additive but not a single code | 18,614 | 38.6 % |
The last row is the important one. In nearly two fifths of lists an additive is present and no E code appears anywhere in the string. Someone scanning for the letter E and not finding it concludes "this one is clean" — and is wrong in more than a third of cases.
Why: the regulation offers a choice
This is not a trick played by particular manufacturers but a direct provision. Part 6 of clause 4.4 of Article 4 of technical regulation TR CU 022/2011 requires the functional purpose of the additive to be stated — acidity regulator, stabiliser, emulsifier — together with its name, "which may be replaced by the additive index under the International Numbering System (INS) or the European numbering system (E)". Text of TR CU 022/2011, page 9.
So the name is the default and the code is a permitted substitute. Writing the name instead of the code is not a way around the requirement; it is the ordinary way of meeting it.
In passing, this inverts the usual complaint. "They hide it behind numbers" sounds convincing, but on our data the number is the rarer and the more informative option: with a code you can find the additive in a reference table in a second, with a name not always.
How large the gap is for individual additives
Number of lists out of 48,230 where the substance is named in words, and where it is written as a code.
| Additive | Code | As a name | As a code |
|---|---|---|---|
| Citric acid | E330 | 10,000 | 1,320 |
| Lecithin | E322 | 7,064 | 824 |
| Potassium sorbate | E202 | 4,184 | 916 |
| Ascorbic acid | E300 | 3,929 | 714 |
| Monosodium glutamate | E621 | 2,494 | 1,692 |
| Xanthan gum | E415 | 2,463 | 628 |
| Sodium benzoate | E211 | 2,364 | 656 |
| Carrageenan | E407 | 2,275 | 694 |
| Mono- and diglycerides of fatty acids | E471 | 2,101 | 1,123 |
| Sodium nitrite | E250 | 1,792 | 378 |
| Locust bean gum | E410 | 1,284 | 108 |
| Titanium dioxide | E171 | 254 | 245 |
Citric acid is written as a name seven and a half times more often than as a code. Locust bean gum, almost twelve times. There is one exception, titanium dioxide, where name and code run neck and neck. For the rest the tilt towards the name is overwhelming.
The devices that make an ingredients list look shorter are a separate subject, and swapping a code for a name is only one of them. The others are covered in the article on how manufacturers hide additives.
Which codes appear most often
Among the 11,003 lists that use codes, the top ten look like this.
| Code | What it is | Lists |
|---|---|---|
| E621 | Monosodium glutamate | 1,692 |
| E476 | Polyglycerol polyricinoleate | 1,480 |
| E330 | Citric acid | 1,320 |
| E471 | Mono- and diglycerides of fatty acids | 1,123 |
| E450 | Diphosphates | 1,095 |
| E1422 | Acetylated distarch adipate | 1,077 |
| E331 | Sodium citrates | 1,046 |
| E150 | Caramel colour, all classes from E150a to E150d | 1,032 |
| E220 | Sulphur dioxide | 920 |
| E202 | Potassium sorbate | 916 |
By group the picture is flatter than talk about colourings would suggest: stabilisers and emulsifiers E400–E499 come first (11.2 % of all lists), then antioxidants and acidity regulators E300–E399 (9.4 %), preservatives E200–E299 (7.2 %), and only then colourings E100–E199 (6.4 %).
What the letter itself means and what the first digit stands for is covered in the article on E numbers in the ingredients list, and each individual code is in the food additives reference.
How many additives are in one product
Among lists that use codes the median is 2 codes and the mean 3.5. The distribution is skewed: 2,730 lists carry five or more codes, 872 carry ten or more, and 198 carry fifteen or more. The maximum in the sample is 41 distinct codes in a single card — a cake.
For comparison, the median length of an ingredients list across the whole sample is 10 comma-separated items. A typical product with additives is therefore not "all chemicals" but a dozen ingredients, two of which happen to carry a code.
Where additives are most common
Categories were assigned from the product name, so the boundaries are approximate. The second column is the share of lists where an additive appears in any form; the third is the share where a code specifically appears.
| Category | With an additive | Of those, with a code | Cards |
|---|---|---|---|
| Confectionery | 93 % | 34 % | 7,818 |
| Sausages and meat products | 89 % | 33 % | 2,130 |
| Drinks | 77 % | 21 % | 3,544 |
| Snacks | 67 % | 28 % | 1,064 |
| Dairy and ice cream | 67 % | 23 % | 4,017 |
| Sauces and mayonnaise | 65 % | 25 % | 2,735 |
| Ready-to-cook foods | 47 % | 22 % | 1,071 |
| Bread and baked goods | 45 % | 20 % | 1,431 |
| Canned foods | 42 % | 13 % | 773 |
The gap between the second and third columns holds in every category: the share of products with a code never exceeds 34 %, while the share with an additive reaches 93 %. In no category does scanning for the letter E give an accurate picture.
What follows from this in the shop
Three practical conclusions.
- Counting codes is a poor way to judge an ingredients list. It measures not how many additives there are but how a particular manufacturer habitually writes them.
- The length of the list tells you more. The median is 10 items; a noticeably longer list usually means a multi-component product, and that is visible without decoding every line.
- The functional purpose is mandatory. The regulation requires "colouring", "stabiliser" or "preservative" before the name or the code. Those words find additives faster than the letter E and are harder to miss.
What else to look at in an ingredients list in ten seconds is covered in a separate article.
What this analysis does not show
It says nothing about safety. The share of products containing additives describes labelling and technology, not harm: every code listed above is permitted for use, most have an established acceptable daily intake, and exceeding it through shopping choices is difficult.
Nor does it show the full picture of composition. The regulation does not require substances that carried over with an ingredient and do not affect the properties of the food to be declared — so some additives never reach the string at all. For the same reason 61.4 % remains a lower bound.
And it does not replace checking the pack in your hand: the data comes from product cards, not from the label in front of you. How such databases are built and why they disagree with each other is covered in the article on where nutrition databases get their numbers. The companion analysis of the same shelf, about sugar, is in where there is more sugar than in cola.
Frequently asked questions
- How many products on Russian shelves contain E numbers?
- In our sample of 48,230 ingredient lists from six retail chains, a code of the form E-xxx appeared in 22.8 %, while an additive in any form appeared in 61.4 %. The second figure is a lower bound: flavourings and carry-over substances were not counted.
- Why do labels say "citric acid" instead of E330?
- Because the regulation allows it. Part 6 of clause 4.4 of Article 4 of TR CU 022/2011 requires the functional purpose and the name of the additive, and the name may be replaced by an index under the INS or E system. The name is the default option and the code is the permitted substitute.
- Which E numbers appear most often?
- Among lists using codes the top five are E621 monosodium glutamate (1,692 lists), E476 (1,480), E330 citric acid (1,320), E471 mono- and diglycerides (1,123) and E450 diphosphates (1,095). By group, stabilisers and emulsifiers E400–E499 are the most common.
- Which food categories contain the most additives?
- In our sample an additive in any form appears in 93 % of confectionery, 89 % of sausages and meat products and 77 % of drinks. Canned foods are lowest at 42 %. Categories were assigned from product names, so the boundaries are approximate.
- How many additives does one product usually contain?
- Among lists using codes the median is 2 codes and the mean 3.5. There are 872 lists with ten or more codes, and the maximum in the sample is 41 codes in a single card. The median length of the whole ingredients list is 10 items.
Read next
- How manufacturers hide additives in the ingredient list — A short ingredient list sells better than a long one, and manufacturers know it.
- E-numbers in ingredients: how the code is structured and what it means — An ingredient list contains E330, and it is impossible to tell from this code whether it is citric acid or something better to avoid.
- Where there is more sugar than in cola: an analysis of 34,000 products on Russian shelves — We took an open dataset from Open Food Facts—four and a half million products—and calculated how much sugar is in what is sold in Russia.
- What to look for in an ingredient list in ten seconds — Reading an entire ingredient list at the shelf is impossible: the print is small, and there is no time.
This article is for general information. It is not medical advice, a diagnosis, or a prescription for treatment or a diet, and it does not replace a consultation with your doctor. If you have a health condition, are pregnant, take medication, or follow a diet prescribed to you, decisions about food belong with your doctor.
Figures from regulations, guidelines and studies are given as they stood when this article was prepared and may since have changed; check them against the primary sources. This article is not advertising, an offer, or individual advice, and neither the author nor the site owner is responsible for decisions taken on the basis of it.