Vitamins and minerals: what the Ciqual 2025 table leaves blank, by food group
How much vitamin and mineral data is missing from the Ciqual 2025 table?
In the Ciqual 2025 table (3,484 foods), 41.1% of the cells in the 33 vitamin and mineral columns hold a dash, meaning the content is not known: 29.7% for the 12 minerals and 47.6% for the 21 vitamin columns. Vitamin K2 has a numeric value for only 102 foods out of 3,484, vitamin D for 1,437. A dash does not mean zero: the Anses documentation asks users not to treat it as a zero value.
Source: Anses, Ciqual French food composition table 2025, DOI 10.57745/RDMHWY (version of 19 Nov 2025). Radish calculations on the official Excel file, retrieved 16 Nov 2026. Published .
A "foods rich in iron" ranking built on the Ciqual table can only rank the foods that have a value. Here is the map of the cells without a value, constituent by constituent and group by group, computed on the official Anses file.
What the Anses documentation says about its own gaps
The official documentation (section 3.2.2, in French) states that when a content is not known, a dash stands in place of the value; that users must take missing values into account and not treat them as zeros; and that 30% of mean values are missing across the whole table, 0.1% for alcohol and 95% for vitamin K2. Our recount finds the same extremes (alcohol 0.1%, vitamin K2 95.0%). Across all 74 constituent columns of the Excel file we count 32.3% dashes against 30% in the documentation, which does not state how it counts: we publish our figure with its method rather than reuse theirs.
Which constituents are the least documented?
The table ranks the 33 vitamin and mineral columns by share of dashes across the 3,484 foods. Several columns are forms of the same vitamin (vitamin E and alpha-tocopherol, four folate columns, vitamin D and its D2 and D3 forms): they are counted separately, as in the file. Two pairs are mutually exclusive: no food has a content in both "vitamin E" and "alpha-tocopherol", nor in both "total folates" and "total folates in DFE". A dash in one of them therefore does not mean the information is missing: it may sit in the other. Neither vitamin E column is filled in for 1,155 foods (33.2%), neither total folate column for 1,051 (30.2%).
| Constituent (per 100 g) | Dash (not known) | “traces” | “< threshold” | Numeric value |
|---|---|---|---|---|
| Vitamin K2 | 95.0 % (3,309) | 1.5 % | 0.6 % | 2.9 % |
| Vitamin E | 82.9 % (2,889) | 0.1 % | 0.1 % | 16.8 % |
| Vitamin B9, total folates in DFE | 73.9 % (2,576) | 0.0 % | 0.0 % | 26.1 % |
| Intrinsic folates | 73.8 % (2,572) | 0.6 % | 0.5 % | 25.1 % |
| Vitamin B9, total folates | 56.2 % (1,959) | 0.1 % | 2.6 % | 41.1 % |
| Vitamin D2 (ergocalciferol) | 55.7 % (1,942) | 9.2 % | 9.2 % | 25.8 % |
| Chloride | 55.7 % (1,939) | 0.0 % | 4.4 % | 39.9 % |
| Vitamin K1 | 54.8 % (1,910) | 0.0 % | 8.0 % | 37.2 % |
| Folic acid (fortification) | 53.7 % (1,870) | 0.0 % | 4.5 % | 41.8 % |
| Alpha-tocopherol (vitamin E) | 50.2 % (1,750) | 0.1 % | 4.8 % | 44.9 % |
| Vitamin A activity (retinol equivalents) | 44.8 % (1,562) | 0.0 % | 0.0 % | 55.2 % |
| Beta-carotene | 41.2 % (1,434) | 0.4 % | 11.3 % | 47.1 % |
| Selenium | 39.7 % (1,383) | 0.2 % | 27.6 % | 32.5 % |
| Vitamin D3 (cholecalciferol) | 39.6 % (1,379) | 4.6 % | 13.2 % | 42.6 % |
| Iodine | 39.5 % (1,376) | 0.0 % | 23.5 % | 37.1 % |
| Vitamin B12 | 38.1 % (1,328) | 0.2 % | 3.4 % | 58.3 % |
| Vitamin D | 37.6 % (1,310) | 3.8 % | 17.3 % | 41.2 % |
| Vitamin B5 (pantothenic acid) | 33.3 % (1,159) | 0.1 % | 1.2 % | 65.4 % |
| Manganese | 31.2 % (1,086) | 0.4 % | 9.0 % | 59.4 % |
| Vitamin C | 31.2 % (1,086) | 0.8 % | 16.3 % | 51.8 % |
| Copper | 29.2 % (1,018) | 0.1 % | 9.5 % | 61.1 % |
| Vitamin B2 (riboflavin) | 28.6 % (997) | 0.1 % | 10.1 % | 61.2 % |
| Vitamin B1 (thiamin) | 28.1 % (980) | 0.1 % | 8.5 % | 63.2 % |
| Vitamin B3 (niacin) | 27.4 % (956) | 0.0 % | 9.3 % | 63.2 % |
| Vitamin B6 | 27.2 % (947) | 0.0 % | 5.1 % | 67.7 % |
| Phosphorus | 27.0 % (942) | 0.1 % | 1.5 % | 71.4 % |
| Retinol | 25.8 % (900) | 0.4 % | 10.0 % | 63.8 % |
| Magnesium | 25.8 % (899) | 0.0 % | 0.5 % | 73.7 % |
| Zinc | 25.7 % (897) | 0.5 % | 6.0 % | 67.8 % |
| Potassium | 24.9 % (867) | 0.0 % | 0.4 % | 74.7 % |
| Iron | 23.9 % (831) | 0.2 % | 5.4 % | 70.6 % |
| Calcium | 22.5 % (784) | 0.0 % | 0.7 % | 76.8 % |
| Sodium | 11.5 % (402) | 0.1 % | 4.3 % | 84.1 % |
Reading: vitamin K1 has a dash for 54.8% of foods. "< threshold" = a value below a limit given in the file; "traces" = detected but not precisely quantifiable, or, without analysis, a content estimated as very low but not zero (documentation, section 3.2.3; see also see How to read the Ciqual table). The four columns may not sum to exactly 100% because of rounding.
Do the gaps differ by food group?
Yes, and the spread can be wide for a single constituent. Share of dashes (%) for eleven constituents, by Ciqual food group (groups are the table's own, translated by us; the last column pools all 33 columns).
| Group (foods) | Iron | Iodine | Selenium | Vit. A (RE) | Vit. D | Alpha-tocopherol | Vit. K1 | Vit. C | Folates (DFE) | Total folates | Vit. B12 | Mean of the 33 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Meat, eggs, fish (791) | 18.5 | 42.0 | 38.7 | 59.4 | 39.7 | 54.5 | 60.9 | 40.1 | 98.9 | 37.5 | 26.3 | 46.0 |
| Fruit, vegetables, legumes and oilseeds (653) | 20.2 | 40.7 | 41.2 | 27.9 | 40.7 | 36.4 | 34.8 | 9.6 | 25.0 | 96.2 | 47.6 | 32.1 |
| Starters and composite dishes (407) | 34.6 | 40.0 | 43.7 | 54.1 | 43.2 | 66.3 | 70.3 | 40.8 | 96.6 | 43.5 | 45.7 | 51.2 |
| Sugary products (361) | 26.9 | 38.0 | 37.1 | 38.8 | 34.9 | 42.7 | 49.6 | 32.1 | 74.2 | 54.0 | 37.1 | 40.6 |
| Dairy products (356) | 31.7 | 34.3 | 36.5 | 52.8 | 32.3 | 46.1 | 59.8 | 48.9 | 78.4 | 56.2 | 38.8 | 45.2 |
| Water and other beverages (325) | 22.5 | 32.6 | 35.4 | 32.0 | 32.9 | 70.5 | 68.3 | 20.9 | 76.9 | 45.5 | 33.8 | 34.6 |
| Culinary aids and miscellaneous ingredients (235) | 23.4 | 49.4 | 49.4 | 47.7 | 31.9 | 55.3 | 50.6 | 29.4 | 76.2 | 54.0 | 42.1 | 40.0 |
| Cereal products (214) | 24.3 | 39.3 | 40.2 | 48.6 | 47.2 | 38.8 | 48.1 | 38.8 | 70.1 | 59.8 | 53.3 | 40.5 |
| Fats and oils (72) | 11.1 | 36.1 | 44.4 | 23.6 | 20.8 | 27.8 | 47.2 | 16.7 | 58.3 | 61.1 | 16.7 | 31.0 |
| Infant foods (39) | 0.0 | 25.6 | 10.3 | 28.2 | 2.6 | 28.2 | 59.0 | 7.7 | 100.0 | 2.6 | 2.6 | 26.1 |
| Ice creams and sorbets (30) | 46.7 | 46.7 | 43.3 | 46.7 | 46.7 | 66.7 | 70.0 | 50.0 | 100.0 | 43.3 | 50.0 | 53.5 |
| All foods (3,484) | 23.9 | 39.5 | 39.7 | 44.8 | 37.6 | 50.2 | 54.8 | 31.2 | 73.9 | 56.2 | 38.1 | 41.1 |
- Fruit, vegetables, legumes and oilseeds (653 foods): 32.1% dashes on average, the lowest of the groups with more than 200 foods. Vitamin C is missing for only 9.6% of them, against 50.0% for ice creams and sorbets.
- Starters and composite dishes (407): 51.2%, the highest of the groups with more than 200 foods; meat, eggs, fish (791): 46.0%.
- Two folate columns, two different maps: the dietary folate equivalents (DFE) column has a dash for 25.0% of fruit and vegetables and 98.9% of meat, eggs and fish; the "total folates" column for 96.2% of fruit and vegetables. No food is filled in for both: according to the documentation (section 3.3.9), the "total folates" column is published when the DFE value is not available, without stating its coefficients. The two complement each other but are not on the same scale: do not add them up or compare them directly.
- Infant foods (39): 26.1% on average and 0 dash for iron. Small groups (infant foods, ice creams and sorbets, 30 foods) should be read with care.
Iodine and selenium: about one food in four is "< threshold"
For iodine, 23.5% of foods have a value of the form "< threshold" and 39.5% a dash; for selenium, 27.6% and 39.7%. Sodium is the least incomplete of the list (11.5% dashes).
How to read this table without being misled
- A dash means "content not known in the table", not "the food contains none". The file does not give the reason.
- Adding up values for a food while ignoring dashes understates the total: this is what the documentation asks users to avoid.
- These shares describe the file, not diets: they say nothing about anyone's intake or about what to eat. For nutrition advice, see a health professional.
Method
- File:
Table Ciqual 2025_FR_2025_11_03.xlsx, downloaded on 16 Nov 2026 from the official repository (DOI 10.57745/RDMHWY, version 1.0 of 19 Nov 2025). MD5 checksum 0d9758ce23f3f13dd63a005bc1bb4f2c, identical to the one published by the repository. - Scope: the 3,484 food rows, sheet "composition nutritionnelle". One row ("Dessert (aliment moyen)", code 24999) has no group: it counts in "All foods" but not in the 11 groups (3,483 foods). The 33 columns kept are the 12 minerals (calcium, chloride, copper, iron, iodine, magnesium, manganese, phosphorus, potassium, selenium, sodium, zinc) and the 21 vitamin columns. Salt, energy and macronutrients are excluded.
- Cell classification: a dash ("-"), the word "traces", a value preceded by "<", or a numeric value. No other form exists in these columns; the script stops on any unknown form. For each column the four counts sum to the number of foods.
- Limits: we read the file's mean values (not minimums, maximums, confidence codes or sources). Counts depend on the file version; an Anses update may change them.
- Reproduce: Python (standard library only) on the official file; the script is published with the page: ciqual-2025-lacunes.py (Etalab 2.0 licence; MD5 of the source file given above).
Download the table (CSV)
ciqual-2025-lacunes-vitamines-mineraux-en.csv (396 rows: 33 constituents × 12 levels, all foods and 11 groups). Semicolon separator, decimal point, UTF-8. Columns: INFOODS code, constituent, group, counts (numeric value, traces, "<", dash) and percentages. Group and constituent names are translated by Radish; the French original is in the French CSV.
Licence and citation
The Anses dataset is published under the Etalab Open Licence 2.0 (stated on the dataset record, DOI 10.57745/RDMHWY). The counts and percentages on this page and in the CSV are Radish calculations based on that data, not Anses data; we make them available under the same licence (licence notice shipped with the CSVs). Anses has neither validated nor endorsed these calculations. To cite them:
Radish, based on Anses, Ciqual French food composition table 2025, https://doi.org/10.57745/RDMHWY (version of 19 Nov 2025), calculations of 16 Nov 2026.
Frequently asked questions
How much of the Ciqual 2025 table is missing?
According to the Anses documentation, 30% of mean values. Across the 33 vitamin and mineral columns we count 41.1% dashes.
Which vitamin is the least documented?
Vitamin K2: 102 foods out of 3,484 have a numeric value (95.0% dashes).
Does a dash mean zero?
No. It means the content is not known; the documentation asks users not to treat it as zero.
General information about published data, not nutrition advice.
Ready to try Radish?Private calorie counter, Nutri-Score, no ads. Free.Get Radish →