Evidence-based guide to vitamin B2: FMN and FAD flavoproteins, mitochondrial respiratory chain, mucosal health, migraine prevention, and neon-yellow urine.
Riboflavin rarely becomes the center of attention in discussions about vitamins. It does not have the reputation of an “immune vitamin” like vitamin C, it is not associated with bones as directly as vitamin D, and vitamin B2 deficiency in developed countries is rarely diagnosed as an isolated clinical condition. On multivitamin labels, it often looks like just another line in a long list of B vitamins.
From a biochemical perspective, that impression is misleading.
Riboflavin sits near the center of cellular energy metabolism. The body uses it to produce two key coenzymes — FMN, flavin mononucleotide, and FAD, flavin adenine dinucleotide. These molecules are required by dozens of flavoproteins involved in mitochondrial function, fatty-acid oxidation, antioxidant defense, vitamin B6 metabolism, folate metabolism, conversion of tryptophan into niacin, and many other reactions.
That means vitamin B2 deficiency is not simply a shortage of one more vitamin. Biochemically, it is a situation in which several metabolic systems can begin to function less efficiently at the same time.
But this importance should not be interpreted in the opposite direction either. It does not mean that high-dose riboflavin can “boost mitochondria” in every healthy person. For most people, daily requirements are measured in roughly one milligram, while pharmacological doses in the hundreds of milligrams make sense only in specific clinical situations — the best known today being migraine prevention.
Vitamin B2 is a water-soluble flavin molecule with a characteristic yellow color. But most of its biological role is not carried out by free riboflavin itself.
Once inside the cell, it is converted first into FMN and then into FAD.
In the first step, riboflavin kinase phosphorylates riboflavin using ATP. FAD synthase then adds an adenyl group to produce FAD.
FMN and FAD subsequently serve as cofactors for flavoproteins.
The distinctive feature of the flavin system is its ability to accept and donate electrons in oxidation-reduction reactions. This makes flavins highly effective molecular intermediates between compounds from which cells extract energy and the systems that ultimately use those electrons to generate ATP.
The human flavoproteome includes dozens of FAD- and FMN-dependent proteins. They participate not only in energy metabolism, but also in amino-acid and lipid metabolism, coenzyme Q biosynthesis, antioxidant defense, one-carbon metabolism, nitric-oxide-related signaling pathways, and many other processes.
This is why impaired riboflavin transport or metabolism can produce such a diverse clinical picture.
If one physiological system best illustrates the importance of riboflavin, it is the mitochondrion.
Complex I of the mitochondrial respiratory chain — NADH dehydrogenase — contains FMN. It accepts electrons from NADH and transfers them through iron-sulfur centers to ubiquinone.
Complex II — succinate dehydrogenase — depends on FAD. This system is particularly interesting because it is simultaneously part of the tricarboxylic acid cycle and the respiratory chain: succinate is converted into fumarate, FAD accepts reducing equivalents, and electrons are subsequently transferred into the ubiquinone pool.
FAD is also required by many other metabolic enzymes. Acyl-CoA dehydrogenases, for example, use it during the first steps of mitochondrial β-oxidation of fatty acids.
Therefore, when flavin metabolism is severely disrupted, several energy-producing pathways can be affected at once.
This becomes especially clear not in mild dietary insufficiency, but in rare inherited disorders affecting riboflavin transport or flavoprotein function. In these conditions, biochemical patterns may resemble defects of fatty-acid oxidation or mitochondrial disease, and high-dose riboflavin can sometimes have major therapeutic value.
But the biochemistry needs to be interpreted correctly.
The fact that FAD and FMN are essential to the respiratory chain does not mean that taking extra vitamin B2 will indefinitely increase ATP production in a person who already has adequate riboflavin status. A cofactor is necessary for the system to function, but once physiological requirements are met, increasing the dose does not necessarily increase the output of that system proportionally.
One of the most interesting aspects of vitamin B2 is that its deficiency can interfere with the metabolism of other micronutrients.
Vitamin B6 is a good example.
The biologically active coenzyme form of vitamin B6 is pyridoxal 5'-phosphate, or PLP. One of the key reactions involved in its formation depends on pyridoxine 5'-phosphate oxidase, an FMN-dependent enzyme.
Adequate riboflavin status is therefore required for normal vitamin B6 metabolism. However, saying that vitamin B2 deficiency automatically produces a complete clinical vitamin B6 deficiency would be too categorical. A more accurate statement is that significant riboflavin deficiency can reduce the efficiency of B6-dependent metabolism.
A similar relationship exists with vitamin B3.
The conversion of the amino acid tryptophan into niacin, and subsequently into precursors of NAD, involves FAD-dependent reactions. Severe riboflavin deficiency can therefore interfere with endogenous niacin production as well.
But perhaps the most interesting interaction of vitamin B2 is with folate metabolism.
The enzyme methylenetetrahydrofolate reductase — MTHFR — uses FAD as a cofactor.
Its role is to produce 5-methyltetrahydrofolate, the form of folate required for the remethylation of homocysteine to methionine.
This is where an interesting interaction between genetics and nutrition appears.
The common MTHFR C677T variant produces a less stable form of the enzyme. The effect is particularly evident in people with the 677TT genotype: the enzyme loses FAD more easily, which can reduce its activity and contribute to higher homocysteine concentrations.
In a randomized trial, supplementation with only 1.6 mg of riboflavin per day reduced homocysteine levels in people with the 677TT genotype, whereas the effect was much less pronounced in carriers of other genotypes. Later studies found additional metabolic effects of riboflavin in people with the TT genotype, including effects on one-carbon metabolism and blood pressure.
This is a good example of nutrigenetics: a small change in intake of an ordinary vitamin can have different physiological consequences depending on the structure of a specific enzyme.
But here too, the finding should not be exaggerated.
Having the MTHFR C677T variant does not in itself mean that a person has a disease, nor does it create a universal indication for high-dose vitamin B2. And lowering homocysteine is a biochemical effect that cannot automatically be equated with proven prevention of heart attack, stroke, or other vascular events.
In supplement advertising, vitamins are often described as molecules that directly “neutralize free radicals.”
For vitamin B2, a more interesting mechanism is involved.
FAD is required by glutathione reductase, the enzyme that converts oxidized glutathione, GSSG, back into its reduced form, GSH.
Reduced glutathione then participates in one of the cell’s central antioxidant systems.
The connection between riboflavin and oxidative stress therefore largely operates through maintenance of the enzymatic glutathione system.
When vitamin B2 is deficient, the activity of FAD-dependent glutathione reductase can decrease. This is, incidentally, the principle behind one of the best functional laboratory tests for assessing riboflavin status.
Experimental and clinical data do support a role for riboflavin in antioxidant defense and regulation of lipid peroxidation. However, this does not mean that high-dose vitamin B2 is a universal “antioxidant therapy” for chronic disease. The biochemical mechanism is well established; the clinical usefulness of large doses depends on the specific condition.
Anemia is not considered a specific sign of vitamin B2 deficiency, but the relationship between riboflavin and iron metabolism has become an increasingly interesting area of research.
Riboflavin deficiency can impair mobilization of iron from stores and affect its utilization during erythropoiesis. In several studies, improving vitamin B2 status was accompanied by improvement in hematological parameters, particularly in people who were riboflavin deficient at baseline.
A 2022 review concluded that subclinical vitamin B2 deficiency can indeed contribute to anemia, particularly in women and children, although the relationship among riboflavin, iron metabolism, and erythropoiesis remains complex.
This is another reason not to diagnose vitamin B2 deficiency based on a single symptom.
Cracks at the corners of the mouth, fatigue, anemia, or tongue inflammation can also occur with deficiencies of iron, vitamin B12, folate, vitamin B6, zinc, and a number of other conditions. Isolated ariboflavinosis is much less common than combined nutritional deficiency.
For such a fundamental molecule, the daily requirement is surprisingly small.
U.S. dietary reference values set the RDA for adult men at 1.3 mg per day and for women at 1.1 mg per day. During pregnancy the requirement rises to 1.4 mg per day, and during breastfeeding to 1.6 mg per day.
This illustrates one of the main differences between a vitamin and an energy substrate.
Riboflavin is not consumed in gram quantities like glucose or fatty acids. Its derivatives function inside enzyme systems as cofactors, so relatively small amounts are sufficient to support a large number of biochemical reactions.
Good food sources include liver and other organ meats, milk and yogurt, eggs, meat, almonds, mushrooms, and certain other foods. In many countries, grain products and breakfast cereals are also fortified with riboflavin.
With a varied diet, meeting physiological requirements is usually not difficult.
For example, one cup of milk contains roughly 0.5 mg of riboflavin, a cup of plain yogurt about 0.6 mg, one egg around 0.2 mg, while a small serving of liver can exceed the daily requirement several times over.
Riboflavin has one unusual property that can be observed even outside the laboratory: it is intensely yellow and fluorescent.
The same molecular structure also makes it sensitive to light.
Ultraviolet and visible radiation can degrade riboflavin and its derivatives relatively quickly. This is one reason milk has traditionally been protected from prolonged exposure to strong light in transparent containers.
For the same reason, phototherapy used in newborns with hyperbilirubinemia can increase riboflavin degradation. NIH identifies prolonged light exposure as one of the factors that can reduce vitamin B2 status.
During cooking, another issue is related less to temperature itself and more to the vitamin’s water solubility. During boiling, some riboflavin can leach into the cooking water, so losses may be higher if the liquid is discarded.
The pharmacokinetics of riboflavin are somewhat unusual.
Most dietary vitamin B2 is not ingested as free riboflavin but as FMN and FAD bound within food components. During digestion, these forms are released and converted into compounds that can be absorbed by intestinal cells.
Transport is mediated by specialized proteins in the RFVT family, encoded by SLC52A1, SLC52A2, and SLC52A3.
But this transport system is saturable.
According to NIH, riboflavin from food is generally well absorbed, but active absorption becomes sharply limited with very large single doses. Once a dose exceeds roughly 27 mg at one time, only a relatively small additional fraction is absorbed.
This creates an interesting paradox.
The physiological requirement for an adult is around 1 mg per day. Active absorption saturates at doses in the tens of milligrams. Yet 400 mg per day is used in migraine prevention.
This does not mean that all 400 mg enter the bloodstream.
A pharmacological dose creates a large concentration gradient, so even a limited absorbed fraction can greatly exceed normal dietary exposure. These are the kinds of doses that were historically used in clinical migraine trials.
So the phrase “400 mg of vitamin B2” describes the administered dose, not the amount that ultimately reaches tissues intact.
This is probably the most noticeable effect of riboflavin.
After a large dose, urine can become intensely yellow, lemon-colored, or almost fluorescent.
The explanation is simple.
Riboflavin itself has a strong yellow color and fluoresces. The body’s storage capacity is limited, intestinal absorption is saturable, and excess absorbed vitamin is rapidly excreted by the kidneys.
The change in urine color is therefore an expected consequence of vitamin B2 excretion, not a sign of liver injury or “kidney overload.”
But the opposite conclusion should not be made either.
Bright urine does not prove that “the vitamins were not absorbed at all.” It merely means that some riboflavin is being excreted — which is entirely expected after a dose far above physiological requirements.
Severe ariboflavinosis is relatively uncommon in countries where varied diets are widely available.
In addition, vitamin B2 deficiency rarely occurs in perfect isolation. Someone with significant undernutrition or malabsorption is often simultaneously deficient in iron, folate, vitamin B6, vitamin B12, and other micronutrients.
That makes the symptoms of vitamin B2 deficiency nonspecific.
Classic manifestations include painful cracks at the corners of the mouth, cheilosis, inflammation of the lips, glossitis, and soreness or inflammation of the oral and pharyngeal mucosa. Seborrheic-like skin changes and irritation or redness of the eyes can also occur.
With severe and prolonged deficiency, anemia and lens changes have been described, including a higher risk of cataract-related changes. Classic descriptions of ariboflavinosis also include corneal vascularization.
But neither a “magenta tongue” nor cracks at the corners of the mouth should be considered pathognomonic for vitamin B2 deficiency.
These findings require a broader search for possible causes.
Among the clearest risk factors are monotonous diets low in riboflavin-rich foods and diseases that impair nutrient absorption.
Risk may rise in significant malabsorptive disorders and after certain gastrointestinal surgeries.
Another group includes people who consume little or no dairy or other animal products. A vegan diet can absolutely provide enough vitamin B2, but it requires more deliberate use of plant sources or fortified foods. NIH identifies vegans and people with very low milk intake among groups in whom inadequate riboflavin status occurs more frequently.
Pregnancy and breastfeeding also deserve additional attention, particularly when the diet is restrictive, because physiological requirements are higher.
Alcohol dependence is often accompanied by poor overall nutritional status and can interfere with normal absorption and metabolism of vitamins.
The relationship with thyroid function is also interesting. Older experimental work showed that thyroid hormones influence the conversion of riboflavin into active coenzyme forms. A small study in people with hypothyroidism found reduced activity of FAD-dependent erythrocyte glutathione reductase, which normalized after thyroxine therapy. NIH also lists thyroid hormone insufficiency among possible factors associated with vitamin B2 deficiency.
But this should not be turned into a rule that “all hypothyroidism causes riboflavin deficiency.”
Simply measuring free riboflavin in serum is not the best way to assess long-term status.
Its concentration depends heavily on recent food and supplement intake and can change relatively quickly.
A more informative functional test is EGRAC — the Erythrocyte Glutathione Reductase Activation Coefficient.
The principle is elegant.
Glutathione reductase requires FAD. In the laboratory, enzyme activity is first measured in red blood cells under baseline conditions. FAD is then added and the activity is measured again.
If the enzyme becomes substantially more active after FAD is added, that suggests that it was inadequately supplied with its flavin cofactor inside the body.
Traditionally, an EGRAC value of ≤1.2 is considered consistent with adequate riboflavin status, values of 1.2–1.4 suggest marginal insufficiency, and values >1.4 suggest deficiency.
However, there is an important modern caveat.
NIH notes that the ideal diagnostic cutoffs for EGRAC remain uncertain and that a higher coefficient does not necessarily correspond linearly to the severity of deficiency. The test also cannot be interpreted properly in people with glucose-6-phosphate dehydrogenase deficiency.
Another option is to measure 24-hour urinary riboflavin excretion. In well-nourished adults, urinary excretion is generally at least around 120 micrograms per day, whereas values below 40 micrograms per day have traditionally been used as an indicator of deficiency.
Urinary excretion, however, is much more sensitive to recent food intake and therefore reflects long-term tissue status less reliably.
Most importantly, a healthy person eating a varied diet generally does not need routine EGRAC testing or urinary vitamin B2 measurements. These tests are useful when there are real clinical or nutritional reasons to suspect deficiency.
The difference between a physiological requirement of roughly 1.1–1.3 mg and a pharmacological dose of 400 mg seems enormous.
Yet migraine prevention is one of the best-studied therapeutic uses of high-dose riboflavin.
The mechanistic hypothesis is largely mitochondrial. Some people with migraine show evidence of altered cerebral energy metabolism, so increasing the availability of precursors for FMN and FAD has been proposed as one way of supporting flavin-dependent mitochondrial systems.
However, the popular explanation that 400 mg of riboflavin literally “saturates Complexes I and II, restores phosphocreatine, and raises the threshold for cortical spreading depression” goes beyond what has actually been demonstrated in humans.
The mitochondrial hypothesis is biologically plausible, but clinical efficacy does not tell us that there is one single proven mechanism responsible for the effect.
The best-known trial was published in 1998.
In this small randomized study, 55 adults with migraine received either 400 mg of riboflavin per day or placebo for three months. Migraine frequency and the number of headache days decreased more in the vitamin B2 group. A reduction of at least 50% in headache days was observed in 59% of participants receiving riboflavin versus 15% in the placebo group.
This figure is often used to claim that vitamin B2 “works in about 60% of patients.”
But one small trial should not be transformed into a universal response probability for everyone with migraine.
Later studies produced more heterogeneous results, particularly in children.
Still, the accumulated evidence overall supports possible benefit. A 2026 meta-analysis including 12 clinical trials and 749 participants found reductions in migraine frequency and duration compared with control. Dose-response analysis suggested increasing effects on these outcomes up to doses of 400 mg per day. At the same time, heterogeneity between studies was very high, and reduction in pain intensity was not statistically significant.
Riboflavin can therefore be considered a reasonable option for migraine prevention, but not a universal replacement for pharmacological therapy.
NCCIH notes that some guidelines recommend 400 mg per day, while also characterizing the quality of evidence as limited. Pediatric studies are more inconsistent, so adult protocols should not simply be transferred to children without medical supervision.
Preventive migraine treatments generally cannot be judged by the rule “I took it — did it work or not?”
Riboflavin is no exception.
Clinical studies usually evaluate it over several months, and the classic regimen is 400 mg per day for approximately three months.
So a person who takes vitamin B2 for five days, notices no change, and concludes that it “does not work” is not really reproducing the studied protocol.
A more sensible approach is to track attack frequency and headache days in a diary over several months.
And here it is especially important to distinguish between two very different situations.
In dietary deficiency, the goal is simply to replace what is missing.
In migraine prevention, a dose in the hundreds of milligrams represents a pharmacological use of the nutrient, many times higher than ordinary dietary requirements.
This does not mean that a person with migraine is “deficient in 400 mg of vitamin B2 per day.”
No classic Tolerable Upper Intake Level, or UL, has been established for vitamin B2.
The U.S. Food and Nutrition Board did not set an upper limit because available data have not demonstrated meaningful toxicity even with doses of 400 mg per day used for several months. Limited intestinal absorption and rapid renal excretion of excess vitamin likely contribute to this favorable safety profile.
But the statement “there is no UL” does not mean “any dose is absolutely safe.”
NIH specifically notes that the absence of reported toxicity is not the same as proof that arbitrarily large doses are harmless.
At studied doses, the most common effects are harmless urine discoloration and, in a small minority of people, mild gastrointestinal symptoms. In the classic migraine trial involving 55 participants, only a few mild adverse events were reported, including diarrhea and increased urination.
So the safety profile does appear very favorable.
But taking gram-level doses of vitamin B2 without a specific reason simply because it is water-soluble makes little sense.
Rare inherited defects of riboflavin transport provide a striking example of how powerful a vitamin can become when there is a specific biochemical defect.
Riboflavin transporter deficiency is associated primarily with biallelic pathogenic variants in SLC52A2 and SLC52A3.
The disorder can present with motor and sensory neuropathy, ataxia, weakness, respiratory impairment, optic atrophy, sensorineural hearing loss, and bulbar palsy. These phenotypes were historically described, among other names, as Brown-Vialetto-Van Laere syndrome and Fazio-Londe syndrome.
In this setting, high-dose riboflavin is not a nutraceutical experiment.
GeneReviews recommends starting high-dose therapy as early as possible when the disorder is suspected, even while diagnostic evaluation is ongoing, because treatment can be lifesaving. Doses in the range of 10–50 mg/kg body weight per day are used, and treatment is continued long term once the diagnosis is confirmed.
This is fundamentally different from the situation of a healthy person taking a B-complex supplement.
The high dose is effective here not because “more vitamin makes mitochondria work better,” but because a genetic defect in the transport system can be partially overcome by creating very high substrate availability.
Older biochemical literature contains reports that certain medications can affect enzymes involved in riboflavin metabolism. These mechanistic observations have sometimes been turned into long lists of drugs supposedly capable of causing functional vitamin B2 deficiency.
Such lists may include tricyclic antidepressants, phenothiazines, and other compounds.
But the clinical evidence supporting most of these claims is much weaker than the biochemical literature may suggest.
At present, the NIH Office of Dietary Supplements states directly that riboflavin is not known to have clinically relevant interactions with medications.
Older mechanistic findings remain interesting from a pharmacological perspective, but there is no basis for turning them into a practical rule such as “amitriptyline blocks vitamin B2, so everyone taking it needs riboflavin supplements.”
This is another good example of a general principle in nutrition science: a biochemical interaction in an experiment and a proven clinical need for additional treatment are not the same thing.
In most cases, no.
If the diet is varied and includes dairy products, eggs, meat, fortified grains, or adequate plant sources, meeting physiological riboflavin requirements is usually straightforward.
A separate supplement makes sense when there is a specific reason.
That may include confirmed inadequate intake or malabsorption, a high risk of deficiency due to a severely restricted diet, a clinician-directed indication, or pharmacological high-dose use for migraine prevention.
But the logic “riboflavin participates in mitochondrial metabolism, therefore everyone benefits from 100 mg for energy” does not hold scientifically.
In someone with adequate vitamin B2 status, flavin-dependent enzymes are already supplied with their required cofactors.
Providing one hundred times more precursor does not make the respiratory chain work one hundred times faster.
Riboflavin is a good example of how an unremarkable-looking vitamin can occupy a fundamental position in human physiology.
Vitamin B2 itself is primarily a precursor of FMN and FAD — two coenzymes without which numerous flavoproteins cannot function normally.
Through these molecules, riboflavin becomes connected with mitochondrial electron transport, β-oxidation of fatty acids, the glutathione system, vitamin B6 metabolism, synthesis of niacin from tryptophan, folate metabolism, homocysteine regulation, and erythropoiesis.
But the biochemical centrality of a vitamin does not make megadoses universally beneficial.
For most adults, physiological requirements are only 1.1–1.3 mg per day, and a varied diet can usually meet them.
Deficiency does occur, but it rarely produces one unique symptom. Cracks at the corners of the mouth, glossitis, dermatitis, eye irritation, or anemia should be interpreted in the broader nutritional and clinical context.
When laboratory assessment is needed, EGRAC remains one of the most useful functional markers, although its diagnostic cutoffs are not as absolute as they are sometimes presented.
High-dose vitamin B2 does have specific therapeutic uses. The best known is migraine prevention, where 400 mg per day for approximately three months has a reasonable evidence base and a favorable safety profile, although not everyone responds and the studies remain heterogeneous.
And in rare inherited riboflavin transporter disorders, doses measured in tens of milligrams per kilogram can genuinely be lifesaving.
The main principle is therefore the same as with many other vitamins:
the body needs enough riboflavin, but the maximum dose does not mean the maximum benefit.
Normal biochemistry requires an adequate supply of the cofactor.
A pharmacological dose requires a specific clinical purpose.
Those two situations should not be confused.
This material is for educational purposes only and does not replace medical consultation, diagnosis, or individually prescribed treatment.
Riboflavin is a naturally fluorescent yellow pigment. Unabsorbed amounts are filtered by the kidneys into urine, creating a striking but completely harmless neon-yellow color.
Migraine pathophysiology involves impaired neuronal mitochondrial phosphorylation. Supplemental riboflavin boosts mitochondrial enzyme activity and ATP reserve.