What Happened
Vitamins · Renal Physiology · History of Science · Terminology · Biochemistry Explainer1870s to today · record to 17 Sep 2026
Riboflavin, and what a B-complex tablet does
The Colour Is the Vitamin
Take a 100 mg B-complex tablet and within a few hours the lavatory looks alarming. Nothing is going wrong. The yellow is not a metabolite, not a warning and not a by-product of anything: it is vitamin B2 itself, unchanged, the same pigment the food industry sells as colour E101. What the colour actually tells you is arithmetic. At most 27 mg is absorbed from a single dose; the adult requirement is 1.3 mg a day; and almost nothing is stored. The usual explanation adds a third step — that the kidney actively pumps the surplus out. That step is repeated everywhere, and this record could not open a single source for it.
- 27 mgthe most an adult can absorb from a single dose, whatever the tablet says
- 1.3 mgthe recommended daily allowance for adult men, and 1.1 mg for women
- 0upper safe limits set for riboflavin — because the evidence on adverse effects “is not sufficient” to set one, not because safety was established
- E101riboflavin’s number as a permitted food colouring in Europe. The pigment in the tablet and the pigment in the sweets are the same molecule
Graded by how well each item stands up Sorted by standing. The mechanism and the evidence for the mechanism are separate rows, never one row.
| Standing | What is recorded | Where |
|---|---|---|
| Confirmed | Riboflavin is itself a yellow-orange crystalline solid and a permitted food colouring. Its name is ribose plus flavin, the ring “that imparts the yellow color to the oxidized molecule”, from Latin flavus, yellow | Encyclopaedic record |
| Confirmed | On a well-balanced diet, 60 to 70 per cent of the flavins in urine are riboflavin itself; the rest are metabolites. So most of what colours the urine has not been changed by the body at all | Institute of Medicine, Dietary Reference Intakes (1998), chapter 5 |
| Confirmed | Absorption happens in the proximal small intestine “via a rapid, saturable transport system”, and the ceiling is about 27 mg absorbed per single meal or dose, with an absorption half-life of 1.1 hours. Below that ceiling, roughly 95 per cent of food flavin is taken up | Zempleni and colleagues (1996); the same figure is the reference value in the Dietary Reference Intakes |
| Confirmed | “When riboflavin is absorbed in excess, very little is stored in the body tissues. The excess is excreted, primarily in the urine.” Urinary excretion accounts for about one half of riboflavin’s removal from plasma | Dietary Reference Intakes; Zempleni and colleagues (1996) |
| Confirmed | Recommended daily allowance 1.3 mg for men, 1.1 mg for women. Median intake from food alone is about 2 mg and 1.5 mg; the 95th percentile of intake from food and supplements is 4 to 10 mg a day | Institute of Medicine, Dietary Reference Intakes (1998), chapter 5 |
| Confirmed | No Tolerable Upper Intake Level has been set, and the report says why in its own words: “The evidence on adverse effects is not sufficient to set a Tolerable Upper Intake Level (UL) for riboflavin” | Institute of Medicine, Dietary Reference Intakes (1998), chapter 5 |
| Confirmed | The same report says plainly what that does not mean: the study behind it “was not designed to assess adverse effects. It is possible that chronic administration of these doses would pose some risk” — and, in its summary, “caution may be warranted” | Institute of Medicine, Dietary Reference Intakes (1998), chapter 5 |
| Confirmed | In the randomised trial of 400 mg a day for migraine, three minor adverse events occurred: diarrhoea and polyuria on riboflavin, abdominal cramps on placebo. None was serious | Schoenen, Jacquy and Lenaerts, Neurology 50(2) (1998) |
| Unconfirmed | That the kidney actively secretes riboflavin rather than merely filtering it. It is the third step of the usual explanation and it is repeated everywhere; no source reachable for this article states it. The reference report explains the same observation by limited solubility, limited absorption and rapid urinary excretion, and does not mention secretion at all | Nothing obtained. The papers usually cited could not be read; the one 1970 record retrieved carries no abstract |
| Unconfirmed | The year the pigment was first seen. The specialist history says 1872; the encyclopaedia says 1879 — and cites the specialist history for that sentence. Both name the same chemist and the same observation | Northrop-Clewes and Thurnham (2012) against the encyclopaedic record |
Timeline
The 1870s to 2024 About sixty years passed between seeing the colour and knowing what it was. The first entry is the one two sources cannot agree on.
- 1872 or 1879The English chemist Alexander Wynter Blyth observes a yellow-green fluorescent pigment in milk. The specialist history of the subject dates it to 1872; the encyclopaedic record says 1879, names the substance lactochrome, and cites that same history for the sentence
- Early 1900sThiamin is isolated — the first vitamin — and called B1. Researchers then realise there is at least one more water-soluble factor, and call the remainder the vitamin B-2 complex. It later turns out to be several unrelated compounds
- The 1930sThe growth of young rats on a B2-free diet, and the brightness of the fluorescence, become the assay. The factor is isolated from egg white as ovoflavin and from whey as lactoflavin, and in the United States is also called vitamin G. The encyclopaedic account of these years is internally inconsistent about which came first, and the page does not settle it
- 1934On the encyclopaedic account, the flavins from the different sources are shown to be one and the same compound; the structure is identified, the name riboflavin is settled on, and the vitamin is synthesised
- 1939Riboflavin is shown to be essential in the human diet. It was the second vitamin isolated and the first out of the B-2 complex, and it kept the number
- October 1994An open pilot study gives 49 migraine patients 400 mg of riboflavin daily for at least three months; mean global improvement 68.2%. Its safety sentence has an exception in it: one patient withdrew with gastric intolerance
- January 1996A randomised cross-over study of nine adults, dosed orally at 20, 40 and 60 mg and intravenously at 11.6 mg, fixes the number this whole subject turns on: 27 mg is the most a single dose can absorb
- 1998Two things in the same year. A randomised controlled trial of 400 mg against placebo in 55 migraine patients reports 59% responders against 15%. And the Institute of Medicine publishes the dietary reference intakes, declining to set an upper limit for want of evidence
- 2013A review sets out the riboflavin transporters and their names: three proteins assigned to a new solute-carrier family because they resemble no other, and renamed RFVT1, RFVT2 and RFVT3 (SLC52A1 to A3)
- 202457 countries require wheat or maize flour to be fortified with riboflavin, at 1.3 to 5.75 mg per kilogram; 16 more do it voluntarily. That is a rounding error beside a supplement, which is why a meal does not do what a tablet does
The Argument
Three steps, and one of them has no source here What the colour is, why there is so much of it, and what the missing upper limit does and does not mean.
Start with the thing that surprises people: nothing has happened to the vitamin. Riboflavin in its purified solid form is a yellow-orange crystalline powder, and it is sold as a food colouring under the number E101 — the pigment in the tablet and the pigment tinting a sweet are one molecule. Its name records the fact. Riboflavin is ribose, the sugar in its tail, plus flavin, the ring “that imparts the yellow color to the oxidized molecule”, from the Latin flavus.
And it comes out the same way it went in. On a well-balanced diet, 60 to 70 per cent of the flavins in urine are riboflavin itself; the rest are metabolites, chiefly hydroxymethylriboflavin. So when the colour appears, the great majority of what is producing it has passed through a body that did nothing to it at all. There is even a name for the result: flavinuria.
- E101the same pigment, sold as a food colour
- 60–70%of urinary flavins are riboflavin, unchanged
The second step is why there is so much of it, and it is arithmetic rather than biology. Absorption happens in the proximal small intestine “via a rapid, saturable transport system”, and the ceiling has been measured: about 27 mg from a single meal or dose, with an absorption half-life of 1.1 hours. Below that ceiling the gut is efficient — roughly 95 per cent of food flavin is taken up. Above it, the surplus is not absorbed at all; it carries on into the large intestine and is broken down there by bacteria.
Set that against the requirement. An adult man needs 1.3 mg a day and a woman 1.1 mg; food alone supplies about 2 mg and 1.5 mg. A supplement carrying 50 or 100 mg is therefore doing two things at once: most of it never gets in, and what does get in is far more than the body has any use for. Very little riboflavin is stored, so the surplus leaves, and about half of its removal from the blood is by way of the urine.
- 27 mgthe absorption ceiling for one dose
- about 77×a 100 mg tablet against a man’s daily requirement — this page’s arithmetic, from the two figures above
The third step is where the usual account and the evidence part company. Almost every explanation of flavinuria says the kidney does not merely filter riboflavin but actively secretes it, giving a clearance several times the filtration rate. No source reachable for this article states that. The papers it is usually traced to could not be obtained; the closest record retrieved — a 1970 paper on riboflavin in newborns — carries no abstract at all.
What the reference work does say is different, and sufficient. Asked why such large doses appear to do no harm, the Dietary Reference Intakes answers with three things: riboflavin’s limited solubility, its limited capacity for absorption, and its rapid excretion in the urine. Secretion is not among them. This page therefore carries two steps as mechanism and the third as a claim it could not check, which is not the same as saying it is wrong — only that a reader who wants it should go and find the paper, because this record could not.
- 2steps of the usual explanation that are established here
- 0sources found for the third
One more thing is widely misread, and it is the one that matters for anyone actually swallowing the tablet. There is no Tolerable Upper Intake Level for riboflavin, and that is usually reported as meaning it has been shown to be safe at any dose. The report that declines to set one says the opposite, in its own words: “The evidence on adverse effects is not sufficient to set a Tolerable Upper Intake Level (UL) for riboflavin.” Of the study it leans on, it adds that it “was not designed to assess adverse effects” and that “it is possible that chronic administration of these doses would pose some risk”. Its summary ends “caution may be warranted”.
That is not an alarm. Nothing read here reports harm from riboflavin at ordinary or even at high doses, and the randomised trial at 400 mg a day found three minor adverse events in 55 patients, of which two — diarrhoea and polyuria — were on riboflavin, none serious. The honest summary is that the question has not been studied, and a missing answer has been circulating as a reassuring one.
| The step | What backs it | Standing |
|---|---|---|
| The molecule is the pigment | A yellow-orange crystalline solid, sold as food colour E101, named for its yellow ring; 60 to 70 per cent of urinary flavins are the unchanged vitamin | Established |
| Absorption saturates | A measured ceiling of about 27 mg from a single dose, reported by a randomised cross-over study and adopted as the reference value by the standards body that sets the dietary intakes | Established |
| The kidney secretes it | Nothing that could be opened. The reference work explains the same observation without it, naming limited solubility, limited absorption and rapid urinary excretion instead | Unchecked here |
| It is harmless | No upper limit is set — but expressly for want of evidence, with the report adding that chronic administration at those doses might possibly pose some risk, and that caution may be warranted | Not studied |
What Others Add
Why it took sixty years, and why the numbers have holes The colour was visible from the start. Everything else about it was not.
60 years
Nothing pointed at it
- Beriberi pointed at B1 and pellagra at B3. There is no classical deficiency disease of riboflavin, so nothing in a clinic led anybody towards it
- The tool that worked in the end was “the growth-stimulating properties of the food extracts given to young rats” — and the brightness of the fluorescence tracked the growth, which made the assay quick
- Along the way it was lactochrome, ovoflavin, lactoflavin and, in the United States, vitamin G — four names for one molecule, each from whatever it had just been pulled out of
1872?
A date two sources cannot agree on
- The specialist history of riboflavin opens: the first observation of the fluorescent milk pigment “can be traced to the English chemist Alexander Wynter Blyth in 1872”
- The encyclopaedic article says 1879, names the substance lactochrome — and gives that specialist history as its citation for the sentence
- Neither can be preferred from here, so both are on the page. The same encyclopaedic paragraph also has the 1930s out of order with itself, which is why this article leans on it lightly for that decade
4 · 8 · 10 · 11
The holes are demotions, not omissions
- B4 has been used for three distinct chemicals — choline, adenine and carnitine. Adenine the body makes itself; carnitine is essential for certain worms and not for humans; choline is an essential nutrient but not a vitamin
- B8 was adenosine monophosphate, and is also used for inositol. B11 was a chick growth factor that turned out to be a folate derivative. B10 is disputed: the encyclopaedic record gives para-aminobenzoic acid, a building block of folate
- So B2 kept its number for a reason: it was the first thing pulled out of the B-2 complex and the second vitamin ever isolated. The source for this list carries a needs-more-citations notice, and the page says so rather than presenting it as settled
One place riboflavin is taken at doses nobody would call nutritional is migraine prevention, and it is worth stating because it shows the absorption ceiling from the other side. Two studies from the same Belgian group: an open pilot in 49 patients at 400 mg a day for at least three months, with a mean global improvement of 68.2%; then a randomised trial against placebo in 55 patients, where 59 per cent responded against 15 per cent on placebo, p = 0.002, with a number needed to treat of 2.3.
Four hundred milligrams is roughly fifteen times the 27 mg that a single dose delivers — this page’s own arithmetic from the two figures, and worth holding loosely for that reason. Whatever is doing the work in those trials, most of the tablet is not getting in. And the safety reporting is not as clean as it is usually relayed: the pilot’s much-quoted “no drug-related side effects were reported” begins “with the exception of one patient … who withdrew because of gastric intolerance”, and the randomised trial recorded diarrhoea and polyuria in the riboflavin arm.
- 59% v 15%responders on riboflavin against placebo, in 55 patients
- about 15×the migraine dose against the absorption ceiling — this page’s arithmetic
Conclusion
What to hold, and how loosely The colour is well understood. What it costs, over years, is not.
If somebody asks why their urine has gone neon after a B-complex tablet, the answer is short and solid: the colour is the vitamin. Riboflavin is a yellow pigment, sold as one; the gut can take in about 27 mg from a single dose and the body needs about 1.3 mg a day; almost nothing is stored, and the surplus leaves in the urine still yellow. Nothing has gone wrong, and a meal will not do it — the flour fortification that 57 countries require is measured in milligrams per kilogram of flour, orders below a tablet.
The part to hold loosely is what the tablet costs. The absence of an upper safe limit is not a safety finding; it is a gap. The body that declines to set one says the evidence “is not sufficient”, that the study it leans on was not designed to look for harm, and that chronic dosing at those levels might carry some risk. Nothing read here reports harm. Nothing read here has looked properly, either.
And the third step of the story everyone tells — the kidney actively pumping the surplus out — is on this page as a claim rather than a mechanism, because no source this article could open says it. That may simply mean the paper is behind a door that would not open here.
What is established
Riboflavin is itself a yellow pigment and a permitted food colour; 60 to 70 per cent of urinary flavins are the unchanged vitamin; absorption is saturable and caps at about 27 mg per dose; the daily requirement is 1.3 mg for men and 1.1 mg for women; very little is stored and the surplus leaves in the urine; and no upper intake level has been set.
What is unsettled, and is not the same as settled
Whether the kidney actively secretes riboflavin — widely stated, not sourced here. Whether chronic high doses are safe — the standards body says the evidence is not there either way, and the study behind its position was not designed to look. And the year the pigment was first seen: 1872 in the specialist history, 1879 in the encyclopaedia that cites it.
What this page is not
Medical advice, and not a reason to stop or start anything. It does not establish what else can colour urine, so a change nobody can account for is still worth asking a clinician about. And the two migraine trials are reported, not recommended: both were run under clinical supervision, at doses about fifteen times the ceiling on what a single dose delivers.