Vitamin B12

What vitamin B12 is

Chemically, vitamin B12 is called cobalamin and carries a cobalt atom at its center. It is produced exclusively by microorganisms, not by plants and not by animals. In food, it is therefore found where their products have been stored: in liver, meat, fish, eggs and dairy products. A purely plant-based diet contains practically none.

Two forms are metabolically active: methylcobalamin and 5-deoxyadenosylcobalamin. Supplements usually contain cyanocobalamin, methylcobalamin or hydroxocobalamin, and the cell converts them into one another. That one form is superior to the others has not been shown so far.

How it is supposed to work

Methionine synthase transfers a methyl group from folate to homocysteine, turning it into methionine. If B12 is lacking, homocysteine builds up and folate metabolism stalls. This hits cell division where it runs fast: in the bone marrow. Hence the enlarged red blood cells.

The second enzyme sits in the mitochondria and channels breakdown products of fatty acids and amino acids into the citric acid cycle. Without B12, methylmalonic acid builds up here. This chain is the reason why B12 is marketed as an energy vitamin. The mechanism is cleanly established, but it only says what happens when B12 is lacking. Not that more of it still does something when supply is good.

Correcting a deficiency is not the same as supplementing

Everything is decided at this dividing line. Correcting a deficiency is about restoring a missing function. The intake values also depend on this: EFSA set the Adequate Intake at 4 µg per day, because from this intake, serum cobalamin and holotranscobalamin are within the reference range and methylmalonic acid and homocysteine stay below their cut-offs.

Anyone supplementing with normal levels, by contrast, expects an additional effect. Evidence for this is lacking. In the B vitamin arm, across 14 studies with 27,882 participants, the Cochrane review on cognitively healthy adults arrives at standardized mean differences between minus 0.03 and 0.06.

Measurement matters more than the supplement

If everything depends on whether you are lacking something, measurement is the crux. The standard test, total serum B12, also captures the fraction bound to haptocorrin, which is not available to the cells. Holotranscobalamin measures only the fraction that can be taken up. Methylmalonic acid and homocysteine measure something else: not the amount, but whether metabolism is stalling.

NHANES data were used to test how well the markers discriminate. Against a combined reference indicator, the area under the ROC curve was 0.963 for methylmalonic acid, 0.944 for total B12 and 0.937 for homocysteine. A single serum value in the gray zone is therefore not a diagnosis, but a reason for the second marker.

Who is really at risk

Biochemical deficiency below 148 pmol/l is not common in high-income countries, at around 2 percent. In regions with few animal or fortified foods, it reaches up to 69 percent.

Vegans without a supplement top the list: in a survey of 104 long-term vegetarians, 44.1 percent of vegans were below 200 pg/ml, compared with 15.0 percent of lacto-ovo vegetarians. The German RBVD study in 36 vegans and 36 omnivores, by contrast, found no difference in biomarkers, because almost all of them supplemented.

Impaired absorption adds to this: in old age, food-cobalamin malabsorption, in which B12 is no longer released from food proteins. Metformin, proton pump inhibitors and H2 blockers also interfere here, and after gastric resection, intrinsic factor is missing entirely.

Injection, tablet, sublingual

Many consider the injection more effective. The data do not support this. A Cochrane review with 3 RCTs and 153 participants found no clinically relevant difference from injection at 1,000 µg daily orally; at 2,000 µg, oral administration was even ahead, with a difference of 680 pg/ml. The meta-analysis of 6,098 participants also found no significant difference between oral, sublingual and intramuscular administration.

That high oral doses work is due to the passive absorption pathway. The active pathway via intrinsic factor is saturated at 1 to 2 µg per single dose; of a 1,000 µg dose, only around 1.3 percent is absorbed. This describes what studies have used and is not a recommendation.

What the studies show

Markun and colleagues, Nutrients 2021: the key study for people with adequate status

Meta-analysis and meta-regression of 16 RCTs with 6,276 participants, explicitly in adults without manifest B12 deficiency and without advanced neurological disease. Endpoints: cognitive function, depressive symptoms, idiopathic fatigue. No effect on any cognitive subdomain, no overall effect on depression measures. On fatigue, only a single one of the 16 studies reported data, which is why no analysis was possible.

VITACOG, PLoS ONE 2010: the most interesting positive finding

Double-blind, placebo-controlled single-center trial in 271 people over 70 with mild cognitive impairment; 168 completed the MRI part, 85 on the active treatment and 83 on placebo. Over 24 months, 0.8 mg folic acid, 0.5 mg vitamin B12 and 20 mg vitamin B6 daily. The primary endpoint was the rate of brain atrophy, and it was met: 0.76 percent per year versus 1.08 percent. With baseline homocysteine above 13 µmol/l, the rate was 53 percent lower. The catch: one center, a surrogate endpoint, three vitamins at once.

Cochrane 2017: the surrogate marker does not deliver what it promises

Fifteen RCTs with 71,422 participants, follow-up 1 to 7.3 years. Homocysteine-lowering vitamins versus placebo. For heart attack, no difference, 7.1 versus 6.0 percent, risk ratio 1.02. For all-cause mortality, none either, 11.7 versus 12.3 percent, risk ratio 1.01. For stroke, a small difference in favor of the vitamins, 4.3 versus 5.1 percent, risk ratio 0.90. All with high quality of evidence.

Where the data stop

The data stop precisely at the most common reason for buying. On fatigue without deficiency, there are no robust figures, because in the largest relevant meta-analysis only one of 16 studies collected this endpoint. For cognition and mood in people with adequate status, figures exist, and they are negative. Homocysteine lowering does not translate into fewer heart attacks either.

On the other hand, the basis is thinner than the consensus suggests. The comparison of oral versus intramuscular administration in proven deficiency rests on three studies with a combined 153 participants; clinical signs and quality of life were not collected as an endpoint in any of them. In children, supplementation improves status with high certainty, but for cognition and body height the effects remain at zero. And for the rule of choosing the methyl form in MTHFR variants, there are no randomized comparisons; MTHFR belongs to folate metabolism.

Status, approval and legal

Vitamin B12 is authorized as a nutrient in food supplements and fortified foods; high-dose injection products are medicines. Regulation (EU) No 432/2012 permits five claims, provided the food meets the minimum requirements for a source of B12: contribution to normal energy-yielding metabolism, to normal functioning of the nervous system, to normal psychological function, to the reduction of tiredness and fatigue, and a role in the process of cell division. The sentence on fatigue is legal, but says nothing about an additional benefit with good supply. The reference value in Germany is 4.0 µg daily according to the German Nutrition Society (DGE), in the US 2.4 µg. There is no statutory maximum; the BfR proposes 25 µg per recommended daily intake for food supplements, 6 µg per 100 g for fortified solid foods and 1.6 µg per 100 ml for beverages. Under anti-doping rules, B12 is not a prohibited substance; what is prohibited is the route, because intravenous infusions of more than 100 ml in 12 hours are banned.

Safety

Neither the US Food and Nutrition Board nor the European SCF has set a tolerable upper intake level, citing the low toxicity potential. The BfR derives its proposal of 25 µg from a supplement intake of 100 µg daily at which no adverse effects became known. That does not mean harmless: in the VITAL cohort with 77,118 participants and 808 lung cancers, a ten-year average dose above 55 µg daily from single-nutrient supplements was associated with a hazard ratio of 1.98 in men, more strongly in smokers; not in women and not with multivitamins. This is a cohort, not proof of causation. Otherwise, B12 is considered well tolerated, with rare reports of acneiform reactions and hypersensitivity. Neurological symptoms, suspected pernicious anemia, status after gastric resection and infants of B12-deficient mothers require medical evaluation, because late treatment can leave permanent damage. High doses of folic acid can mask the blood picture of a B12 deficiency while the nerve damage continues.

Sources

• Markun et al., Nutrients 2021 (meta-analysis, 16 RCTs, 6,276 participants)
• Martí-Carvajal et al., Cochrane Database Syst Rev 2017, CD006612
• Wang et al., Cochrane Database Syst Rev 2018, CD004655 (oral versus intramuscular)
• Rutjes et al., Cochrane Database Syst Rev 2018, CD011906
• Larvie et al., Cochrane Database Syst Rev 2026, CD015264 (children)
• Smith et al., PLoS ONE 2010 (VITACOG)
• Mineva et al., Curr Dev Nutr 2026 (diagnostics: MMA, B12, homocysteine)
• Brasky et al., J Clin Oncol 2017 (VITAL cohort, lung cancer)
• EFSA NDA Panel, Dietary Reference Values for cobalamin, EFSA Journal 2015
• BfR, proposed maximum levels for vitamin B12 in foods including food supplements