Methylcobalamin vs Cyanocobalamin: Which Form of Vitamin B12 to Choose, According to Studies

★ OUR FORMULA

B Vitamins

All 8 B vitamins combined in a single vegetable capsule, with vitamin B12 in methylcobalamine form (2.5 µg), vitamin B9 in active folate form (400 µg of 5-methyltetrahydrofolate), and vitamin B6 in pyridoxal-5-phosphate form (5 mg). One capsule per day, in the morning at breakfast, for a course of 1 to 3 months.

View B Vitamins

The term vitamin B12 designates a family of molecules, the cobalamins, built around a cobalt atom. What differentiates them is the group attached to this cobalt: a cyano group for cyanocobalamin, a hydroxyl group or a water molecule forhydroxocobalamin, a methyl for methylcobalamin and an adenosyl foradenosylcobalamin. A 2015 review published in Molecular Nutrition and Food Research reminds us that the methyl bond is among the strongest in this family, and the water bond among the weakest.

Two of these forms work directly in the body. Methylcobalamin serves as a cofactor for methionine synthase, which recycles homocysteine into methionine; adenosylcobalamin serves as a cofactor for methylmalonyl-CoA mutase, in the mitochondria. Hence the name "active forms". According to a 2023 review devoted to forms of vitamin B12, methylcobalamin dominates in plasma, adenosylcobalamin in tissues, and an adult's body contains altogether 2 to 3 mg of vitamin B12, about half of which in the liver.

Cyanocobalamin, on the other hand, barely exists in nature: it was the first to be isolated, from liver tissue, because the purification method fixed a cyano group onto the vitamin, and it is only found in trace amounts in tissues, related to cyanide introduced notably by tobacco. It nevertheless dominates in medications and supplements, because it is the most stable. Hydroxocobalamin is the predominant form in foods and represents about half of the vitamin B12 bound to transcobalamin in non-supplemented individuals.

The most widespread argument in favor of methylcobalamin can be summed up in one sentence: it would already be active, therefore directly usable, whereas cyanocobalamin must first be converted. The biochemistry is more nuanced. All forms travel in the blood bound to the same transport protein, transcobalamin, and enter the cell by the same mechanism. Inside, the same protein, called MMACHC, removes the group attached to the cobalt, whether it is cyano, methyl, or adenosyl.

The 2015 review draws a clear conclusion from this: methylcobalamin and adenosylcobalamin most likely follow the same intracellular pathway as cyanocobalamin, and their supplementation has little chance of offering an advantage over it. The authors nevertheless recognize the value of high doses of injected hydroxocobalamin, 1 to 2 mg, in certain hereditary diseases of vitamin B12 metabolism.

A 2017 review, though favorable to natural forms, reaches the same conclusion: all forms are reduced to a common core, subsequently converted to methylcobalamin and adenosylcobalamin in a proportion that does not depend on the form ingested, and the methyl as well as the adenosyl provided by the supplement are detached without serving to manufacture the new coenzymes. These authors prefer natural forms, but estimate that for the majority of the population, all probably have comparable effects. A 2015 editorial recalled in turn that the two coenzymes have distinct roles and proposed treating a deficiency through their combination, by hydroxocobalamin, or by cyanocobalamin.

The differences therefore bear more on the destination than on the speed. In cultured human cells, a Danish team compared hydroxocobalamin and cyanocobalamin in 2024: ultimately, both were taken up at over 90% and resulted in the same distribution, approximately 40% methylcobalamin, 40% adenosylcobalamin, and 20% hydroxocobalamin. But hydroxocobalamin was processed six times faster and accumulated twice as fast, a possible sign of a slightly earlier response, to be confirmed in humans.

To be absorbed, vitamin B12 from food is released from its proteins by stomach acidity, taken over by a protein in saliva and then by intrinsic factor, secreted by the stomach, before being captured at the end of the small intestine. This system is very efficient but saturates quickly: the 2015 review mentions a capacity of about 1.5 µg per meal. A second pathway, passive diffusion, allows about 1 to 2% of the swallowed vitamin to pass through without intrinsic factor; it becomes essential when the latter is lacking, after stomach removal or in Biermer's anemia, and it is this pathway that allows high oral doses to correct a deficiency.

The percentages are those from the 2023 review; the absorbed amounts are calculated by us based on these percentages. A 2013 systematic review on vitamin B12 from food points in the same direction: absorption varied from 4.5% for 38 µg supplied by liver to 83% for 3 µg supplied by mutton meat.

What about the form? The 2015 review brings together older data obtained in humans with labeled vitamin. After an oral dose of 1 µg, the retained portion ranged from 34% for adenosylcobalamin to 56% for hydroxocobalamin; at 5 µg, from 13% for adenosylcobalamin to 20% for cyanocobalamin; at 25 µg, from 6% for cyanocobalamin to 8% for adenosylcobalamin. The authors conclude that these studies show no consistent and clear difference between forms, neither for absorption, nor for elimination, nor for clinical effects. A more recent study also places saturation between 3 and 6 µg per dose, above the 1 to 2 µg per meal figure often cited.

Once absorbed, vitamin B12 must still be retained. According to the 2015 review, injected hydroxocobalamin at 200 µg remains longer in the plasma and passes less into urine than an equivalent dose of cyanocobalamin, with strong individual variations. The 2023 review also reports, for cyanocobalamin, lower tissue retention and stronger urinary excretion, while considering the bioavailability of the four forms comparable for most people.

This study illustrates a frequent pitfall: blood levels inform us about what circulates, not what the organs have taken up. Experiments in rats by the same team point in this direction. After a single dose, hydroxocobalamin accumulated more in the liver, cyanocobalamin in the brain, muscles and plasma. In rats deficient then fed one or the other form for two weeks, cyanocobalamin raised plasma B12 by 131%, versus 68%, but more than half remained in the form of cyanocobalamin in the kidneys, and it provided less than half of the active coenzymes obtained with hydroxocobalamin.

For methylcobalamin, data are more sparse. A 2025 narrative review cites animal studies where urinary excretion of cyanocobalamin was three times higher than that of methylcobalamin, with 13% more cobalamins stored in the liver for methylcobalamin, and recommends prioritizing the latter. These animal results remain an indication: no human trial has linked this retention to better clinical outcomes.

Head-to-head trials are rare and converge toward equivalence when the dose is sufficient. The most robust compares two forms and two routes after gastric bypass, surgery that eliminates a large part of normal absorption.

In children, two retrospective Turkish studies directly compared the forms. In 129 young people aged 5 to 18 years who were deficient, injected cyanocobalamin, sublingual cyanocobalamin and sublingual methylcobalamin all corrected vitamin B12 levels and blood abnormalities, even though the magnitude of increase differed between groups. In 158 children aged 0 to 3 years, sublingual methylcobalamin proved as effective as oral or injected cyanocobalamin in raising levels above 300 ng/L. These studies are not randomized: each child received the form chosen by their doctor.

The most recent meta-analysis on routes of administration, published in 2025, brings together 16 studies and 6,098 participants: 14 used cyanocobalamin, 6 methylcobalamin and 4 hydroxocobalamin. The authors do not report analysis by form and find no difference between routes, with very high heterogeneity.

Among the studies we identified, no randomized trial in adults compared oral methylcobalamin to oral cyanocobalamin at equal doses and on clinical criteria. Available trials show that each form tested corrects a deficiency when the dose is appropriate.

In the absence of large trials against cyanocobalamin, methylcobalamin was tested alone, against placebo or at different doses, particularly in diabetic neuropathy, nerve damage that causes tingling, burning, and loss of sensation in the feet.

Doubling the dose does not necessarily bring more benefit. In a 16-week Iranian trial of 35 diabetic patients with low B12, of whom 32 completed it, 1,000 and 2,000 µg of methylcobalamin daily improved symptoms comparably; the higher dose raised blood B12 more without additional benefit, and was accompanied by a significant decrease in kidney filtration rate, observed in a small population. Considering all B vitamins combined, a 2026 meta-analysis (13 trials, 834 patients) presents a mixed picture: symptom questionnaire and sural nerve conduction improved, but the clinical examination score (−0.39 point) remained below the clinically important threshold of 0.5 point, and pain measured on a numeric or visual scale did not differ significantly.

Following total stomach removal for cancer, a 2026 review identified seven studies and 449 patients, of whom 310 were treated orally, most often with 500 to 1,500 µg of methylcobalamin daily: the oral route was not inferior to injections on blood markers, but only one study was randomized and follow-up rarely exceeded six months. In amyotrophic lateral sclerosis, a Japanese trial of 130 patients used 50 mg of injected methylcobalamin twice weekly for 16 weeks and slowed functional decline by 1.97 points compared to placebo. This specialized treatment tells us nothing about oral methylcobalamin.

Cyanocobalamin releases its cyano group in the cell when it is transformed. But each vitamin molecule, very large, carries only one cyano group: the amount involved remains very low at nutritional doses, and none of the identified studies reports cyanocobalamin toxicity taken at usual doses by a healthy person.

The debate concerns two populations. For smokers, already exposed to tobacco cyanide, some researchers propose natural forms rather than cyanocobalamin for prolonged supplementation. In test tubes, cigarette smoke extracts seem to even transform methylcobalamin and hydroxocobalamin into cyanocobalamin, a result to verify in the body. For people whose kidneys function poorly, the main argument comes from a Canadian trial.

One of the authors, J. David Spence, a specialist in stroke prevention, argues that high-dose cyanocobalamin causes cyanide accumulation in case of kidney insufficiency, and that B vitamins would be beneficial when kidneys function well but harmful when filtration drops below 50 mL/min; he considers it likely that methylcobalamin should then replace cyanocobalamin. He presents thiocyanate accumulation, the cyanide derivative eliminated by the kidneys, as a possible explanation, recommends methylcobalamin or hydroxocobalamin in affected patients, and calls for studies to settle the matter.

The link between cobalamin and cyanide also works in reverse: hydroxocobalamin, which binds cyanide very strongly, serves as an antidote in case of poisoning and increases its elimination by the kidneys.

All cobalamins degrade under the effect of light, heat, humidity, extreme pH, and certain compounds like vitamin C, and the 2023 review classifies cyanocobalamin as the most stable. Under ambient light, the bond carrying the methyl or adenosyl group breaks, and methylcobalamin like adenosylcobalamin in solution transform into hydroxocobalamin, a form still usable by the body.

The differences are significant. In a study cited by the 2023 review, in solution under ultraviolet A at physiological pH, hydroxocobalamin was the most stable, cyanocobalamin degraded approximately five times faster, and methylcobalamin like adenosylcobalamin in seconds. In solid products, six commercial methylcobalamin preparations in tablets and capsules lost 5 to 31% of their methylcobalamin in their original packaging after 20 days of light exposure, and 8 to 44% once opened, depending on packaging and coating.

The route often matters more than the form. In a randomized trial published in 1998 in the journal Blood, 33 evaluable deficient patients received either 2 mg of cyanocobalamin daily by mouth for 120 days, or 1 mg by injection nine times over three months: correction of blood and neurological abnormalities was equally rapid in both groups, and the average vitamin B12 level at four months reached 1,005 pg/mL with tablets versus 325 pg/mL with injections.

A 2018 systematic review published in the Cochrane Database of Systematic Reviews included three randomized trials and 153 participants followed for three to four months. With 1,000 µg daily by mouth, two trials found no clinically relevant difference with injections; with 2,000 µg daily, the 1998 trial found a higher level of 680 pg/mL favoring tablets. Oral treatment was less expensive, but the quality of evidence remained low.

This one-year result is one of the rare signals unfavorable to the oral route: in maintenance phase, a weekly 1 mg tablet performed slightly worse than monthly injection. The 2025 meta-analysis finds no difference between oral, sublingual, and injectable routes, but with very high heterogeneity and signs of publication bias.

The route sublingual does not perform better than the swallowed tablet. In an Israeli trial, 30 deficient patients received 500 µg of cobalamin daily under the tongue, the same dose swallowed, or a B-complex vitamin: in four weeks, the level rose from 94 to 288, from 108 to 286, and from 98 to 293 pmol/L, with no difference between groups. In 40 vegetarians and vegans with mild deficiency, 50 µg daily of sublingual cyanocobalamin improved markers in 12 weeks with no difference from 2,000 µg taken once weekly.

A clinical practice guideline published in 2014 in the British Journal of Haematology summarizes hematologists' position: oral treatment can be appropriate provided doses are taken appropriately and regularly, and treatment should not be delayed when clinical signs are clear, to avoid neurological damage.

Before choosing, you need to know where you stand. Blood vitamin B12 testing remains the first test, but it lacks precision; a 2021 review on diagnosis emphasizes the value of holotranscobalamin, methylmalonic acid, and homocysteine to confirm early deficiency or an uncertain result.

The 2015 review proposes a guideline according to the cause. Insufficient dietary intake, in a vegan for example, is prevented with oral supplements up to 10 µg daily; poor absorption of vitamin B12 from food, common with age, with oral doses below 0.5 mg daily; Biermer's anemia requires at least 0.5 mg daily by mouth, or at least 1 mg by injection at regular intervals.

The dose needed to correct deficiency can be surprising. In a Dutch trial of 120 elderly people with mild deficiency, 16 weeks of oral cyanocobalamin at 2.5, 100, 250, 500, or 1,000 µg daily reduced methylmalonic acid by 16, 16, 23, 33, and 33%: it took 647 to 1,032 µg daily to achieve 80 to 90% of maximum effect. In this trial, the 2.5 µg dose was cyanocobalamin; our B Vitamins provide 2.5 µg of methylcobalamin. The amount is the same, the form differs: a complex at nutritional dose accompanies daily intake, without replacing prescribed treatment for deficiency.

For daily maintenance, our B Vitamins combine methylcobalamin, vitamin B9 in the active form of folate, and six other B vitamins in one vegetable capsule, to be taken in the morning at breakfast with a large glass of water, in a 1 to 3 month course that can be renewed. Our article on the proven effects of B vitamins provides details on the complex.

Frequently asked questions

Sources

Learn more