Spotting a vitamin B12 deficiency: serum, holo-TC, MMA
Futures Nutrition Editorial Team · 12 August 2026
Spotting a vitamin B12 deficiency: serum, holo-TC, MMA
The short answer: a single vitamin B12 value in serum does not answer the question. It is the cheapest of the four available markers, the one measured most often — and the one that changes last. The German Nutrition Society (DGE) puts it unambiguously in its questions-and-answers paper on vitamin B12: to assess supply, a meaningful functional marker, for example methylmalonic acid, should be determined together with a status marker — that is, serum vitamin B12 or holo-transcobalamin.
Two values instead of one, and from two different categories. That is the whole point of this text. Anyone who leaves the practice with "my B12 was normal" has a snapshot of transport in the blood, not a statement about whether enough is arriving inside the cells.
Why the total value says so little
Vitamin B12 does not circulate freely in the blood; it is bound to two transport proteins. By far the largest share — 80 to 90 % — hangs on haptocorrin and is biologically inactive; it is not taken up by cells. Only the remaining 10 to 20 % sits on transcobalamin, and precisely this share is called holo-transcobalamin, or holo-TC for short. It is the form that reaches the cell via its own receptors.
The standard laboratory value "vitamin B12" measures both together. It is therefore determined predominantly by the share that does not matter. Two things follow that regularly cause confusion in practice: the total value falls only late, when the reserves have long been shrinking — and it can even be high while holo-TC is low, for instance when more haptocorrin is being produced.
On top of that comes a broad grey zone. Values below 160 pg/ml indicate a deficiency with high specificity; values above 350 to 400 pg/ml largely rule it out. In between lies a large zone in which the finding is simply open. How big a difference the choice of cut-off makes is shown by a figure from laboratory medicine practice: applying the old threshold of 200 ng/l in older people, 10 to 15 % count as undersupplied. Take the 200 to 400 ng/l range or the finer parameters instead, and the share rises to as much as 50 %.
The four markers compared
In its derivation of dietary reference values for cobalamin, the EFSA names exactly four markers: serum cobalamin, holo-TC, methylmalonic acid (MMA) and total homocysteine. And it adds a sentence that gets to the heart of the matter: the sensitivity and specificity of these markers can be affected by factors unrelated to B12 status — the limitations of all markers make a combination necessary.
| Marker | What it measures | Common decision threshold | What distorts it |
|---|---|---|---|
| Total vitamin B12 (serum) | total transport, predominantly inactively bound | below 160 pg/ml suggests a deficiency, above 350–400 pg/ml argues against one | haptocorrin, wide grey zone, biotin in the assay |
| Holo-transcobalamin (holo-TC) | the share that reaches the cells | below 40 pmol/l suggests a deficiency | impaired kidney function, biotin in the assay |
| Methylmalonic acid (MMA) | whether too little is actually arriving in the cell | above 32 µg/l in serum suggests a deficiency | impaired kidney function, bacterial overgrowth in the gut |
| Homocysteine | the same metabolism, but non-specifically | laboratory-dependent | folate and B6 supply, kidney function, sample transport |
The first two are status parameters: they describe how much is there. The last two are functional parameters: they rise because a B12-dependent enzyme cannot do its work, and thus describe a state inside the cell. That is why combining one marker from each group produces a picture that no single value can deliver.
One detail about homocysteine that is easily overlooked: after the blood draw the value continues to rise artificially, by around 10 % per hour, because the red blood cells keep releasing homocysteine. Without a stabilised special tube, what you end up measuring is the transport time to the laboratory.
The order in which things change
A deficiency does not develop in a day. Laboratory medicine describes it in stages, and from those stages it becomes clear why the choice of marker decides the result.
| Normal | Stage I | Stage II | Stage III | with anaemia | |
|---|---|---|---|---|---|
| Holo-TC (pmol/l) | above 50 | below 35 | below 35 | below 35 | below 35 |
| Homocysteine | normal | normal | normal | elevated | elevated |
| MMA | normal | normal | normal | elevated | elevated |
| MCV (size of the red blood cells) | normal | normal | normal | normal | elevated |
| Haemoglobin | normal | normal | normal | normal | reduced |
Stage I is the early negative balance, stage II the emptying of the stores, stage III affects blood formation. The most striking point of this table: holo-TC is already reduced in the first stage, while blood count and functional parameters are still entirely unremarkable. An isolated low holo-TC shows that the stores are emptying; if elevated methylmalonic acid is added, the deficiency has already reached the metabolism.
The full blood count is therefore not suitable as a screening test. Megaloblastic anaemia is indeed the most common clinical manifestation of a B12 deficiency, but it stands at the end of the chain — and the neurological abnormalities occur independently of it, often even before. A normal MCV rules nothing out.
Why the total value is nevertheless usually the one measured is not a technical question but an economic one. Under German statutory-insurance billing, total vitamin B12 costs around €4, holo-TC just under €15, homocysteine about €14 and methylmalonic acid €43. The cheapest test is the one that answers the question worst.
What falsifies the value — including from your own cupboard
Three confounders are so widespread that they belong in every assessment.
Kidney function. Methylmalonic acid is excreted exclusively via the kidneys. Even a slight impairment — the rule rather than the exception in older people — raises the serum value without any lack of B12. The way out is measurement in the second morning urine, referenced to creatinine; a value that is still falsely high after that comes practically only from bacterial overgrowth of the gut.
Your own intake. The total value in serum responds to the last tablet, not to the store. Anyone who supplements for three weeks and then has a measurement taken is measuring the tablet. The right moment therefore comes before starting an intake — not after it.
Biotin. This is the point hardly anyone has on their radar, even though it comes straight off the supplement shelf. Many laboratory assays work with a streptavidin-biotin system; excess biotin in the blood interferes with them. For the German Federal Institute for Drugs and Medical Devices (BfArM) this warranted a Dear Healthcare Professional Communication ("Rote-Hand-Brief") in 2019: depending on the assay design, values come out falsely elevated (competitive assay principle) or falsely reduced (sandwich principle). For the common Roche assays this means, concretely for our topic — vitamin B12 is measured falsely high from a biotin concentration of 60 ng/ml, holo-TC falsely low from 40 ng/ml, folic acid falsely high from 20 ng/ml. Three values, three directions, one finding.
Such concentrations do not arise from food but from preparations: laboratories ask to be informed when 10 mg of biotin a day or more is being taken — or a smaller amount within the last twelve hours. 10 mg is exactly the 10,000 µg printed on many skin-hair-nails products. Anyone taking something like that is better off saying so before the blood draw than being pleased afterwards about a reassuringly high B12 value.
How far below this figure the documented threshold lies is shown by the procedure from which the Rote-Hand-Brief emerged. In 2019 the Pharmacovigilance Risk Assessment Committee (PRAC) of the European Medicines Agency assessed the available data and found sufficient evidence of interference with clinical laboratory tests from an oral intake of 150 µg biotin per day (for infusions from 60 µg); below that, the data were not sufficient for a conclusive assessment. Medicines containing 150 µg or more per dosage unit have carried a warning in the package leaflet ever since. This obligation does not apply automatically to food supplements — they are foods, not medicines. The German Federal Institute for Risk Assessment (BfR) recommends the note anyway: biotin-containing food supplements should as a matter of principle state that before a laboratory test one should inform the doctor or the laboratory staff about the intake. Incidentally, the ng/ml figures above describe the concentration in the blood, the 150 µg the daily intake — two different quantities, connected only through metabolism.
To put the amounts in context: the EFSA has derived an adequate intake of 40 µg biotin per day for adults, the D-A-CH societies give 30 to 60 µg, and the German National Nutrition Survey II measured a median of between 39 and 48 µg per day in adult men and women. One tablet of biotin 10,000 µg — we too carry this strength common in the trade — therefore delivers around 250 times the adequate intake and about 67 times those 150 µg from which the data on assay interference are considered sufficient.
This is expressly not a question of tolerability: no tolerable upper intake level has ever been derived for biotin, because in 2001 the SCF lacked the systematic dose-response studies for it, and the BfR still considers a maximum level for food supplements dispensable, since no adverse effects have been observed even far above the adequate intake. It is a question of timing. Particular attention applies with impaired kidney function, in children and in pregnancy: biotin is excreted via the kidneys, the half-life lengthens accordingly, and the risk of interference rises with the dose (BfArM/PEI, Bulletin zur Arzneimittelsicherheit 4/2018).
Not among the confounders, but in the same "looks like supply, isn't" category, are the vitamin B12 analogues from algae, sauerkraut and similar products. The DGE is clear about them: they do not contribute to an adequate supply and can even worsen supply further by blocking the transport systems.
An elevated homocysteine is not yet a B12 diagnosis
Homocysteine is the second functional parameter alongside methylmalonic acid, costs less than a third — and answers a different question than most people assume. The reason lies in the metabolism itself: homocysteine sits at a junction with two exits. On one it is remethylated to methionine; for this, methionine synthase needs a methyl group from 5-methyltetrahydrofolate and methylcobalamin as a coenzyme. On the other it is broken down via cystathionine, and the two enzymes responsible work with vitamin B6. A backlog therefore arises from a gap in folate just as much as from a gap in B12 or in B6 — the value rises without saying which of the three substances is missing. Add to that kidney function, which plays into this just as it does with methylmalonic acid, and transport to the laboratory, which pushes the value up by around 10 % per hour.
That is exactly why homocysteine appears in the table above as "non-specific". As sole evidence of a B12 deficiency it is not suitable; as confirmation alongside a low holo-TC it is usable, and methylmalonic acid is the more specific of the two functional parameters because it depends on a single B12-dependent enzyme only.
Why folic acid can change the finding
The connection between folate and B12 reaches beyond the laboratory value, and it has a direction that matters for the assessment. Because both substances are needed in the same reaction, the blood count looks the same with a B12 gap and with a folate gap: the red blood cells become too large. A generous folic acid intake can normalise this blood count again while the B12 gap remains unchanged — the conspicuous finding disappears, the cause does not.
This consideration is not a footnote but the endpoint from which the upper limit for folic acid was derived: in 2000 the EU Scientific Committee on Food took the possible masking of a vitamin B12 deficiency from an intake of 5 mg folic acid per day as its basis, divided by an uncertainty factor of 5 and thus arrived at a tolerable upper intake level of 1 mg per day for adults. The EFSA confirmed this value in 2023 after re-examination (1,000 µg for adults and pregnant women, 800 µg for adolescents aged 15 to 17). It applies to synthetic folic acid and other authorised folate sources from preparations and fortified foods — not to folate from a normal diet.
| Vitamin B12 | Folic acid | |
|---|---|---|
| Nutrient reference value on the label | 2.5 µg | 200 µg |
| Reference value for adults (D-A-CH) | 4.0 µg | 300 µg folate equivalents |
| Tolerable upper intake level | not derived | 1,000 µg (EFSA 2023) |
| BfR proposed maximum per daily portion | 25 µg | 200 µg |
The BfR proposal of 200 µg per recommended daily portion comes from the February 2024 update (opinion 009/2024). Independently of that, the institute expressly maintains its recommendation of 400 µg folic acid daily for women of childbearing age and for pregnant women in the first trimester — there it is about preventing neural tube defects, that is, a different question from the maximum level on the shelf.
For measurement this means two things. First: anyone supplementing folic acid should mention it when a B12 finding is being assessed — not because the test would be falsified as with biotin, but because the blood count drops out as an indicator. Second: the B12 question belongs with a marker of its own anyway, and that is the argument for holo-TC or methylmalonic acid instead of a full blood count.
Our folic acid 1000 µg contains pteroylmonoglutamic acid, the synthetic and longest-studied form, 180 tablets at one a day. The 1,000 µg correspond exactly to the tolerable upper intake level for adults — no room is left on top of that for fortified foods, and they lie well above the BfR proposal of 200 µg per daily portion. Further strengths can be found in the folic acid category.
Authorised for folate is, among others, this claim: Folate contributes to normal blood formation. (Regulation (EU) No 432/2012). It applies to folate alone — there is no joint authorised claim for folate and vitamin B12, even though the two work in the same reaction step.
Who should have a measurement taken
The store is what makes vitamin B12 a special case among the water-soluble vitamins. An adult carries around 2.5 mg, of which roughly 1 mg is in the liver, and gives up only 0.1 to 0.2 % of it per day — regardless of how full the store currently is. This reserve covers requirements for three to four years. An abrupt stop of intake therefore only shows up years later, and that is precisely why a blood value is more sensible than waiting for symptoms in certain constellations. How much is actually needed, and why tablets lie so far above that, is covered in our article on the daily vitamin B12 requirement.
A measurement is indicated, among others:
- with a purely plant-based diet — the DGE recommends that people who eat exclusively plant-based have their supply checked medically on a regular basis;
- from around the age of 65, because gastric acid production declines and B12 is then released less well from dietary protein;
- with diseases of the gastrointestinal tract — atrophic gastritis, chronic inflammatory bowel diseases, exocrine pancreatic insufficiency or after removal of parts of the stomach or intestine;
- with long-term use of certain medicines: proton pump inhibitors, H2 receptor blockers, cholestyramine, various antibiotics and metformin inhibit absorption in the gut.
Once a deficiency is established, the question becomes its cause. Antibodies against intrinsic factor are highly specific for pernicious anaemia but capture only around 40 % of cases; antibodies against parietal cells are more sensitive (up to 90 %) but also occur in about 20 % of people with other autoimmune conditions. Together the two tests find roughly 96 % of those affected. That is diagnostics in medical hands — not a shelf topic.
What the value does not decide
A laboratory value does not replace clinical assessment, and that holds in both directions. With clear symptoms, a therapeutic trial is considered indicated even when the biomarkers cannot demonstrate a deficiency. Conversely, the EFSA describes a state for which there are no symptoms at all: cobalamin insufficiency — biochemical abnormalities such as elevated homocysteine or elevated methylmalonic acid without any specific sign of illness.
Anyone supplementing after a measurement will find the strengths and forms side by side in the vitamin B12 category; the tablet most common in the trade — ours included, vitamin B12 1,000 µg — delivers 1,000 µg as methylcobalamin.
Which form is printed on the label changes nothing about the measurement — how cyano-, methyl-, hydroxo- and adenosylcobalamin actually differ is set out in vitamin B12 forms compared. Because homocysteine also depends on folate supply, both are often looked at together when the value is abnormal; folic acid preparations therefore almost always stand alongside in consultations. Why a B12 gap can be overlooked even as the blood count improves is explained in vitamin B12 and folic acid.
Authorised for vitamin B12 are, among others, these claims: Vitamin B12 contributes to normal red blood cell formation. Vitamin B12 contributes to normal functioning of the nervous system. Vitamin B12 contributes to the reduction of tiredness and fatigue. (Regulation (EU) No 432/2012). Food supplements are not a substitute for a balanced and varied diet and a healthy lifestyle.
Frequently asked questions
My B12 value was normal — can I still have a deficiency? Yes. The total value predominantly measures the inactive fraction bound to haptocorrin and falls only late. With increased haptocorrin production it can even be high while holo-TC is already low. That is why the DGE recommends combining a status and a functional marker.
Which value is the best? There is no single one. Holo-TC changes earliest and is superior to the total value in sensitivity and specificity; methylmalonic acid is the only marker that shows whether too little is actually arriving in the cell. The EFSA expressly considers the combination of several markers necessary, because each one has its limits.
Should I leave out the preparation before the blood draw? The sensible time for measurement is before starting an intake, because otherwise the status parameters reflect the last tablet. If the intake is already running, it belongs in the medical history — as does a high-dose biotin preparation, which can shift the B12 value upwards and holo-TC downwards in the assay.
My homocysteine is elevated — do I have a B12 deficiency? Not necessarily. Homocysteine backs up as soon as one of three routes jams: with too little folate, with too little vitamin B12 or with too little vitamin B6. Add to that kidney function and the transport of the sample to the laboratory. The value shows that something is not running smoothly, not what. More specific for vitamin B12 is methylmalonic acid, because it depends on a single B12-dependent enzyme only.
I take biotin — do I have to stop it before the blood draw? That is decided by the practice or the laboratory, not by the packaging. Either way it should be mentioned: in 2019 the EMA saw sufficient evidence of interference with laboratory tests from 150 µg biotin per day, and usual skin-hair-nails preparations lie far above that at 10,000 µg. Laboratories usually ask for an interval of at least twelve hours; with impaired kidney function more may be necessary, because biotin is excreted more slowly. Affected are not only B12 and holo-TC but many immunoassays — including thyroid and cardiac markers.
How long does it take for a change to show up in the blood? A long time. The body store of around 2.5 mg lasts three to four years if supply was previously good, and it is emptied by only 0.1 to 0.2 % per day. Signs of deficiency therefore usually appear only after years — with one important exception: in people without a previously filled store, such as breastfed infants of vegan mothers without an assured B12 supply, this can happen considerably sooner.
