Vitamin B12 (Cobalamin)
vitamin · also known as cobalamin, cyanocobalamin, methylcobalamin, mecobalamin, hydroxocobalamin, adenosylcobalamin, dibencozide, vitamin B-12, B12
This is general information, not medical advice. Supplements can interact with medicines and health conditions. Always check with your doctor or pharmacist before starting or changing supplements.
What it is
Vitamin B12 is a water-soluble vitamin required for red blood cell formation, DNA synthesis and the maintenance of the myelin sheath around nerves. It is the B vitamin where the gap between who benefits and who does not is widest. In people who are genuinely deficient, replacing it works: deficiency causes a macrocytic anaemia and a progressive neurological damage that can become permanent if left long enough, and treatment reliably restores blood levels. In people who are not deficient, the evidence that extra B12 improves energy, mood or thinking is weak to absent, despite that being the reason most people buy it. What makes B12 worth paying attention to is who quietly becomes deficient: it occurs naturally almost only in animal foods, so unsupplemented vegans are at real risk; absorption depends on stomach acid and on intrinsic factor, both of which decline with age and with long-term acid-suppressing medication; and metformin measurably lowers it over years of use. Because the liver stores B12, a shortfall develops silently over years rather than weeks, which is precisely what makes it easy to miss.
Which form to choose
Grades here rate how well the evidence supports choosing a form, which is not the same as how much that form has been studied: a well-studied form can be graded low precisely because the studies argue against it. This is a different scale from the claim grades below — see our editorial policy.
The synthetic form used in most supplements, most fortified foods and most of the clinical trial evidence, including the trial behind the oral-versus-injection claim below. It is the most chemically stable form and the best studied. The body converts it to the active coenzyme forms, releasing a trace of cyanide so small it is irrelevant at supplement doses in people with normal kidney function. High oral doses work not because absorption is efficient but because it is not: the intrinsic-factor pathway saturates at roughly 1.5-2 mcg per dose, and the remainder crosses the gut wall by passive diffusion at a rate of about 1%. That is why a 1,000 mcg tablet is a sensible dose and a 5 mcg one is not, for anyone whose absorption is impaired.
Graded on trial use. It is the form in most supplements, most fortified food and most of the trial evidence, including the randomised non-inferiority trial of oral against intramuscular treatment that this monograph's main claim rests on, and the systematic review of oral and nasal treatment. Held at B rather than A because the question those trials answer is the route, oral against injection, rather than this form against another one: no trial establishes cyanocobalamin as superior to its alternatives, only as the one that was studied.
One of the two coenzyme forms the body actually uses, and widely sold on the premise that this makes it superior to cyanocobalamin. No head-to-head randomised trial comparing the two forms for correcting deficiency or improving any clinical outcome appears in the published literature, and the major clinical review of B12 deficiency treats cyanocobalamin as the standard without identifying methylcobalamin as better. Methylcobalamin does have its own modest evidence base in diabetic peripheral neuropathy specifically, which is a different question from whether it is a better general-purpose B12. Buy it if you prefer it; do not pay a premium expecting it to work better, because that comparison has not been made.
Graded on the absence of a comparison rather than on a negative result. No head-to-head randomised trial comparing methylcobalamin with cyanocobalamin for correcting deficiency or for any clinical outcome appears in the published literature, and the systematic review of oral B12 treatment works from cyanocobalamin throughout without identifying this form as better. It sits above a bare D because it does have its own evidence base in diabetic peripheral neuropathy, which is a real but different question from whether it is the better general-purpose B12, and below cyanocobalamin because the treatment evidence is cyanocobalamin's.
- Adenosylcobalamin (dibencozide)
The other coenzyme form, active in mitochondrial metabolism, sold on much the same "more active" premise as methylcobalamin. We give no dosage protocol for it because we could find no human trial establishing one. That is a statement about the evidence, not a judgement that the form does not work — but a dose we cannot cite is a dose we will not print.
We have not yet graded the evidence for choosing this form.
- Hydroxocobalamin
Retained in tissues longer than cyanocobalamin, which is why it is the form usually chosen for intramuscular injection, given by a clinician on a schedule of loading doses followed by maintenance every one to three months. It is not a self-administered supplement and we give no dosage protocol for it here deliberately. Injected B12 is the right treatment where absorption is severely impaired — pernicious anaemia, or after gastric surgery — and that is a conversation with a doctor, not a purchase.
We have not yet graded the evidence for choosing this form.
Evidence-based benefits
In adults with confirmed B12 deficiency, high-dose oral B12 normalises blood levels about as well as intramuscular injection over the first two months, though injection may hold a small advantage by one year.
The OB12 trial randomised 283 primary-care patients aged 65 and over with B12 deficiency to oral cyanocobalamin (1 mg daily for 8 weeks, then 1 mg weekly) or intramuscular B12. At 8 weeks more than 90% of both groups had normalised serum B12, with a per-protocol difference between them of just -0.7% (95% CI -3.2 to 1.8), meeting the trial's non-inferiority threshold. By 52 weeks a gap had opened — 73.6% normalised on oral against 80.4% on injection, a difference of -6.3% (95% CI -11.9 to -0.1) — and the Bayesian probability that oral remained non-inferior at one year was low (0.036-0.060). Side effects and quality of life were comparable, and 83.4% of patients preferred taking tablets. A systematic review of oral and nasal B12 in people whose deficiency stems from gastrointestinal disease reached a similar conclusion. The grade is B rather than A because this rests largely on one pragmatic trial rather than several of the same design. The practical reading: tablets are a reasonable first choice for most people, and injection remains preferable where absorption is severely impaired or where adherence over years is uncertain.
In people who are not deficient, B12 supplementation does not produce a clinically meaningful improvement in cognitive function, and there is no reliable evidence that it improves energy or reduces fatigue.
A Cochrane review pooled 14 randomised controlled trials (27,882 participants, mostly over 60) comparing folic acid, B12, B6 or their combination against placebo, and found little or no effect on global cognitive function at any point up to 5 years, with standardised mean differences between -0.03 and 0.06, and -0.01 at 5-10 years. Certainty was rated moderate to low. The literature is not perfectly consistent: a smaller meta-analysis of 23 B-vitamin trials reported a small but statistically significant benefit for global cognition (SMD -0.18, 95% CI -0.30 to -0.06) and episodic memory (SMD -0.09, 95% CI -0.15 to -0.04). We weight the Cochrane analysis more heavily on size and method, but the disagreement is real and worth knowing about. Two honest limits apply to both: neither isolates B12 from co-administered folate and B6, and neither was designed to test the "energy" claim that sells most B12 in the first place.
- Grade A· Strong evidence[1]
Vitamin B12 added to folic acid produces a further reduction in blood homocysteine beyond what folic acid achieves alone, though lowering homocysteine has not been shown to reduce cardiovascular events.
A meta-analysis of individual participant data from 12 randomised controlled trials (1,114 people) found folic acid at 0.5-5 mg/day lowered blood homocysteine by about 25% at typical Western baseline levels, and that adding vitamin B12 at around 0.5 mg/day produced a further 7% reduction (95% CI 3-10%). Vitamin B6 added nothing significant. Reductions were larger in people starting with higher homocysteine or lower folate. This is a well-established effect on a blood marker, and the claim is deliberately scoped to that marker: large outcome trials since have generally not shown that lowering homocysteine with B vitamins reduces heart attacks or overall cardiovascular death. A biomarker moving in the right direction is not the same as a person being better off, and the distinction matters here more than most places.
Risks & contraindications
Long-term metformin treatment roughly doubles the risk of biochemical B12 deficiency and measurably lowers serum B12, with the effect growing with dose and duration.
A meta-analysis of 31 studies found metformin users had about double the risk of B12 deficiency compared with non-users (risk ratio 2.09, 95% CI 1.49-2.93) and significantly lower serum B12 (mean difference -63.70 pmol/L, 95% CI -74.35 to -53.05), a mean relative fall of 14.68% from baseline, scaling with both dose and treatment duration. That analysis did not find a significant association with anaemia or neuropathy prevalence specifically. An earlier meta-analysis reached the same directional conclusion. The grade is C because this is pooled observational data rather than trials designed to test it, though the mechanism — metformin interfering with calcium-dependent uptake of the B12-intrinsic factor complex in the ileum — is well characterised and the association is large and consistent. If you take metformin long-term, this is a reason to have your B12 checked rather than to assume a supplement has covered it, particularly because folic acid in a combined product can mask the anaemia while nerve damage continues.
- Grade C· Preliminary evidence[1]
Adults following an unsupplemented vegan diet have materially worse functional B12 status than omnivores, with vegetarians at intermediate risk; supplementation corrects it.
A meta-analysis of 17 studies found vegan adults had significantly lower serum B12 than omnivores (standardised mean difference -0.72, 95% CI -1.26 to -0.18) and significantly higher total homocysteine (SMD 0.57, 95% CI 0.26-0.89), with a non-significant trend toward higher methylmalonic acid (SMD 0.28, 95% CI -0.01 to 0.57). That combination matters: raised homocysteine and methylmalonic acid indicate a genuine functional shortfall rather than merely a lower reading on a reference range. Vegans also differed from vegetarians on serum B12 (SMD -0.25, 95% CI -0.40 to -0.10) and homocysteine (SMD 0.24, 95% CI 0.09-0.39). Subgroup analysis found supplementation significantly improved every biomarker among vegans. The grade is C because these are cross-sectional and cohort comparisons, not trials — but the finding is unusually actionable, because the cause is dietary source rather than metabolism, and it is fully correctable with a cheap tablet.
- Grade C· Preliminary evidence[1]
Long-term use of high-dose single-ingredient B12 supplements is associated with roughly double the risk of lung cancer in men, particularly smokers; the association was absent in women and absent for multivitamins, and an observational cohort cannot establish cause.
The Vitamins and Lifestyle (VITAL) cohort followed 77,118 adults aged 50-76 and identified 808 incident lung cancers through cancer registry linkage, using 10-year average daily dose as the exposure. B12 and B6 taken as individual supplements — but not the same vitamins in a multivitamin — were associated with a 30-40% increase in lung cancer risk in men, with no association in women. At the highest 10-year average doses the risk in men approached double: hazard ratio 1.98 (95% CI 1.32-2.97) for B12 above 55 mcg/day, and 1.82 (95% CI 1.25-2.65) for B6 above 20 mg/day. Risk was higher among men smoking at baseline and was present across histological types except adenocarcinoma, the type least tied to smoking. This is one observational cohort, and the pattern could reflect something about who takes high-dose single B vitamins rather than the vitamins themselves, which is why it is graded C. It is included because the doses concerned are ordinary — 55 mcg/day is far below a standard 1,000 mcg tablet — and because the authors' own conclusion was that these supplements are not chemopreventive and may be harmful.
Dosage & how to take it
- Cyanocobalamin: 1000 mcg/day, any time of day, with or without food. 1,000 mcg/day is the oral treatment dose from the OB12 trial, which normalised serum B12 in over 90% of deficient patients within 8 weeks before dropping to 1,000 mcg weekly for maintenance. It is also the dose most commonly sold. It looks absurd against a reference intake of 2.4 mcg/day, and the reason is absorption rather than need: the intrinsic-factor pathway saturates at roughly 1.5-2 mcg per dose, so a large oral dose is a way of harvesting the ~1% that crosses the gut wall passively. Someone eating a normal mixed diet with healthy absorption does not need this and gets enough from food. Someone supplementing preventively — a vegan, say, with no absorption problem — needs far less than a treatment dose, and a lower-strength daily tablet or a weekly one is reasonable. No tolerable upper intake level has been set for B12, reflecting low toxicity at oral doses; the lung cancer signal above is a long-term-use caution for men on high-dose single supplements, not a toxicity ceiling. No duration limit is set because no evidence establishes one, and for people with a permanent absorption problem, treatment is lifelong by design.
- Methylcobalamin: 1000–2000 mcg/day, any time of day, with or without food. The range comes from a 2026 randomised trial in patients with diabetic peripheral neuropathy and low B12, which compared 1,000 mcg against 2,000 mcg of oral methylcobalamin over 16 weeks and found the two equally effective for neuropathic symptoms; the higher dose raised serum B12 further without further symptom benefit, and was associated with a small decline in estimated kidney filtration rate. On that evidence there is no reason to reach for 2,000 mcg by default. Nothing indicates methylcobalamin needs a different dose from cyanocobalamin to achieve the same correction in blood levels, so treat the two ranges as interchangeable until a head-to-head trial says otherwise — and see the note on this form above before paying more for it.
Interactions
Metformin interferes with the absorption of vitamin B12 in the final section of the small intestine, and over years of use this measurably drains the body's B12. The evidence here is unusually strong for a supplement interaction: a 4.3-year randomised placebo-controlled trial found metformin lowered B12 levels by 19% on average and produced roughly one extra case of outright deficiency for every 14 people treated, and 13 years of follow-up in the Diabetes Prevention Program found the risk of deficiency grew with each additional year of metformin use, with nerve symptoms more common in metformin users whose B12 was low. A meta-analysis of 31 studies confirms about double the risk of B12 deficiency in metformin users, with the fall in B12 depending on both the dose and the duration of treatment. Because the liver stores B12, the shortfall develops silently over years rather than weeks. Taking a B12 supplement is a sensible safeguard here, and this is one of the clearest cases on the site where supplementing is the helpful move rather than the risky one. Two caveats. Metformin impairs absorption, so a standard oral dose relies on the small fraction that crosses the gut wall passively — which is an argument for a proper 1,000 mcg tablet rather than a token amount buried in a multivitamin. And if you take folic acid alongside it, whether separately or in a B-complex, the folate can correct the anaemia of B12 deficiency while nerve damage quietly continues, so supplementing is not a substitute for testing. If you take metformin long-term, ask your prescriber to check your B12 level periodically — the authors of the trial, the Diabetes Prevention Program follow-up, and the meta-analysis all recommend exactly this.
Stomach acid is needed to release vitamin B12 from the protein in food, so medicines like omeprazole that switch acid production off can, over years, reduce how much B12 you absorb from meals. A large US case-control study comparing nearly 26,000 people with newly diagnosed B12 deficiency against 184,000 without it found that two or more years of PPI use was associated with 65% higher odds of deficiency, and higher daily doses carried more risk; a systematic review and meta-analysis of the available studies reached the same conclusion, with roughly a 1.8-fold increased risk from long-term acid suppression. These are observational findings, so they show an association rather than proof, and it is worth knowing that a more recent meta-analysis found a weaker pooled effect and judged the association less clear-cut. The mechanism, though, points the remedy at supplements rather than away from them: the B12 in a supplement is not bound to food protein, so it does not depend on stomach acid to be freed in the same way. If you take a PPI long-term, a B12 supplement is a reasonable safeguard rather than a risk — and ask your prescriber to check your B12 level rather than guessing, especially if you notice persistent fatigue, numbness, or tingling.
These two work on the same pathway, and taking folate without knowing your B12 status is the one combination worth thinking about carefully. Folate and B12 both feed the conversion of homocysteine to methionine, and a shortage of either produces the same macrocytic anaemia. Folic acid can correct that anaemia on its own — which sounds like a good thing, and is the problem. The anaemia is the visible warning sign of B12 deficiency; the damage that matters is neurological, and it continues underneath while the blood count looks normal. In an analysis of 1,459 older Americans after folic acid fortification began, people with low B12 who also had high serum folate had roughly three times the odds of anaemia and 2.6 times the odds of cognitive impairment compared with those with low folate, while in people whose B12 was normal high folate was, if anything, protective. The harm is specific to the pairing of high folate with low B12, not to folate itself. Taking the two together is therefore the sensible arrangement rather than a risky one, which is why this is marked supportive: adequate B12 removes the thing that makes high folate a problem. Two caveats worth carrying. A supplement is not a test — if you have a reason to suspect low B12, such as a vegan diet, being over 60, or long-term metformin or acid-suppressing medication, ask for it to be measured rather than assuming a combined tablet has covered it. And the evidence is genuinely contested: a 2024 review judges the masking concern compelling enough to warrant vigilance while noting it rests largely on correlational data, and other authors argue it is overstated. The practical advice is the same either way.
Our medication coverage is not exhaustive. Absence of a warning here is not evidence of safety. Always confirm with your pharmacist.
Where to buy
Checking current prices…