Supplement encyclopedia

Vitamin B12

Vitamin B12 is needed for nervous system function and the formation of red blood cells. It comes from animal foods and fortified products. Supplements may contain cyanocobalamin or methylcobalamin, for example.

Double-check the conclusions and review the sources before choosing a supplement.

Main conclusion

Potential benefits

  • A B12 supplement helps raise blood levels of the vitamin. This was shown in vegans and vegetarians with low B12 after three months of taking tablets under the tongue.
  • In people aged 75 and over with moderately low B12, without anemia or pronounced memory impairment, tablets taken for a year showed no clear improvement in memory compared with placebo.
  • In adults with fatigue associated with irritable bowel syndrome or inflammatory bowel disease and B12 levels of at least 150 pmol/L, supplementation for eight weeks did not reduce fatigue more than placebo.
  • Methylcobalamin may slightly increase peak power in a short cycling sprint. In male recreational cyclists who trained, peak power was about 4% higher than with placebo after three days of supplementation.
  • B12 may reduce pain from diabetic nerve damage. In people with type 2 diabetes and B12 below 400 pmol/L who had been taking metformin long term, supplementation added to usual treatment improved the pain score over a year compared with placebo.

[1] [2] [3] [4] [5]

What was studied

  • In people with low B12, researchers studied how supplementation raises blood levels of the vitamin. Memory and fatigue were assessed separately in people with different starting conditions.
  • Short cycling sprints in men who trained and symptoms of nerve damage in people with diabetes were also studied. You can focus on the result for the goal that matters to you.

[1] [2] [3] [4] [5]

Forms studied

  • Supplements contain cyanocobalamin or methylcobalamin, including tablets to swallow or place under the tongue. Check the amount of B12 and how to take it on the label.
  • In people aged 65 and over with low B12, cyanocobalamin was studied at 1 mg per dose: taken orally every day for the first eight weeks, then once a week; injections every other day for the first two weeks, weekly through week eight, then monthly. For blood B12 levels, oral treatment met the “no worse than injections” test at eight weeks, but not at one year.
  • Vegans and vegetarians with low B12 took cyanocobalamin under the tongue; researchers compared two doses. Consider the starting vitamin level when choosing a regimen.

[1] [2] [3]

How to get B12 from food

  • B12 is found in mollusks, fish, meat, eggs, and dairy products. For example, 170 g of plain nonfat yogurt provides about 1.04 micrograms, and a boiled egg weighing 50 g about 0.56 micrograms.
  • On a vegan diet, fortified foods or a supplement help provide B12. Check the label for the amount of the vitamin in a particular product.

[1] [2]

Effects and studied questions

How to read the certainty scale 1–4

Certainty is rated separately for each outcome. 1 is very low, 2 is low, 3 is moderate and 4 is high.

Goal or questionResultConfidenceStudiesPopulation studiedKey sources
Fatigue in people with bowel conditionsFatigue decreased by 8.1 points with B12 and 8.3 points with placebo on the CIS subjective fatigue subscale. B12 provided no additional reduction in fatigue.
2/4 LowOne study with 95 participants. The available description provides insufficient detail to fully verify allocation concealment or how the reported results were selected. The assessment applies to fatigue in this group and cannot be generalized to all causes of fatigue.
11 studySurplus vitamin B12 use does not reduce fatigue in patients with Irritable Bowel Syndrome or inflammatory bowel disease: A randomized double-blind placebo-controlled trial ↗95 adults aged 18–65 with marked fatigue and either irritable bowel syndrome or inflammatory bowel disease. Baseline B12 was ≥150 pmol/L.
  1. Surplus vitamin B12 use does not reduce fatigue in patients with Irritable Bowel Syndrome or inflammatory bowel disease: A randomized double-blind placebo-controlled trial ↗
Memory in people aged 75 and older with moderately low B12On a word-list memory test, the B12 group scored an average of 1.4 words lower than the placebo group after adjustment for baseline scores. The 95% confidence interval ranged from −2.9 to +0.1 words. The study found no advantage for B12 over placebo in memory performance or on the other cognitive tests studied.
2/4 LowOne study. Cognitive results were available for 184 of 201 participants. The interval for the estimate includes the possibility of no difference. The result applies to older adults without pronounced symptoms of deficiency.
11 studyEffects of vitamin B-12 supplementation on neurologic and cognitive function in older people: a randomized controlled trial ↗201 people aged 75 and older with B12 levels of 107–<210 pmol/L at screening, without anemia or marked memory impairment. Cognitive data were available for 184 participants.
  1. Effects of vitamin B-12 supplementation on neurologic and cognitive function in older people: a randomized controlled trial ↗
Peak power in a short cycling sprintThe highest peak power recorded across five 30-second sprints was 764.1 W with B12 and 734.2 W with placebo, a difference of 29.9 W, or about 4.1% (p=0.016). The study did not assess endurance over longer efforts.
2/4 LowThe study included only 18 men and three days of use per condition. It did not test whether the result would persist with longer-term training. Participants received both treatments in sequence. The seven-day break leaves open the possibility of a carryover effect between periods.
11 studyShort-Term Methylcobalamin Supplementation Is Associated with Changes in Anaerobic and Cognitive Performance in Amateur Cyclists: A Randomized Crossover Trial ↗18 male recreational cyclists aged 18–45 who were training. Each completed both study conditions.
  1. Short-Term Methylcobalamin Supplementation Is Associated with Changes in Anaerobic and Cognitive Performance in Amateur Cyclists: A Randomized Crossover Trial ↗

The study count refers to the cited source. Reviews with overlapping trials and outcomes from one study are not added together.

Daily intake requirements and upper limits

Daily B12 needs from food and supplements combined
GroupB12 per dayWhat the reference value means
Infants aged 0–6 months0.4 µgAdequate Intake (AI), total from food and supplements
Infants aged 7–12 months0.5 µgAdequate Intake (AI), total from food and supplements
Children aged 1–3 years0.9 µgRecommended Dietary Allowance (RDA), total from food and supplements
Children aged 4–8 years1.2 µgRecommended Dietary Allowance (RDA), total from food and supplements
Children aged 9–13 years1.8 µgRecommended Dietary Allowance (RDA), total from food and supplements
Adolescents aged 14–18 years2.4 µgRecommended Dietary Allowance (RDA), total from food and supplements
Adults aged 19 and older2.4 µgRecommended Dietary Allowance (RDA), total from food and supplements
Pregnancy, age 14 and older2.6 µgRecommended Dietary Allowance (RDA), total from food and supplements
Breastfeeding, age 14 and older2.8 µgRecommended Dietary Allowance (RDA), total from food and supplements

Vitamin B12 — Health Professional Fact Sheet (NIH ODS) ↗

Forms

Cyanocobalamin and methylcobalamin: The studies used both forms of B12. There are no data showing that one is better absorbed than the other. The key details on the label are the amount of B12 per dose and how to take it. Swallowing B12 or letting it dissolve under the tongue: Sublingual tablets showed no overall advantage over swallowed B12. The study in vegans and vegetarians compared two doses of sublingual cyanocobalamin, not two routes of administration. Comparison with injections: In the study of people aged 65 and older, participants took cyanocobalamin solution by mouth daily for eight weeks, then weekly. At one year, this regimen could no longer be confidently considered comparable to injections. If you have impaired absorption or pronounced symptoms, choose how to replace B12 with your doctor. B12 with no chemical form specified: The fatigue study described oral B12, while the study of borderline B12 levels described tablets. The available descriptions of these studies did not specify the chemical form.

Guidance for each goal

Fatigue in people with bowel conditions2/4 LowRead moreCollapse
Confidence
2/4 LowOne study with 95 participants. The available description provides insufficient detail to fully verify allocation concealment or how the reported results were selected. The assessment applies to fatigue in this group and cannot be generalized to all causes of fatigue.

There was no additional relief from fatigue.

Study findings. Fatigue decreased by 8.1 points with B12 and 8.3 points with placebo on the CIS subjective fatigue subscale. B12 provided no additional reduction in fatigue.

Studied regimen1,000 µg per dayOral B12
Duration and population8 weeks95 adults aged 18–65 with marked fatigue and either irritable bowel syndrome or inflammatory bowel disease. Baseline B12 was ≥150 pmol/L.

Comparator: Placebo

Memory in people aged 75 and older with moderately low B122/4 LowRead moreCollapse
Confidence
2/4 LowOne study. Cognitive results were available for 184 of 201 participants. The interval for the estimate includes the possibility of no difference. The result applies to older adults without pronounced symptoms of deficiency.

No advantage in memory performance was observed.

Study findings. On a word-list memory test, the B12 group scored an average of 1.4 words lower than the placebo group after adjustment for baseline scores. The 95% confidence interval ranged from −2.9 to +0.1 words. The study found no advantage for B12 over placebo in memory performance or on the other cognitive tests studied.

Studied regimen1 mg (1,000 µg) per dayCyanocobalamin, oral tablets
Duration and population12 months201 people aged 75 and older with B12 levels of 107–<210 pmol/L at screening, without anemia or marked memory impairment. Cognitive data were available for 184 participants.

Comparator: Identical-looking placebo tablets

Peak power in a short cycling sprint2/4 LowRead moreCollapse
Confidence
2/4 LowThe study included only 18 men and three days of use per condition. It did not test whether the result would persist with longer-term training. Participants received both treatments in sequence. The seven-day break leaves open the possibility of a carryover effect between periods.

Peak power was about 4% higher in the short test.

Study findings. The highest peak power recorded across five 30-second sprints was 764.1 W with B12 and 734.2 W with placebo, a difference of 29.9 W, or about 4.1% (p=0.016). The study did not assess endurance over longer efforts.

Studied regimen1 mg (1,000 µg) per dayMecobalActive methylcobalamin, sublingual tablets
Duration and population3 days per condition, with a 7-day break between conditions18 male recreational cyclists aged 18–45 who were training. Each completed both study conditions.

Comparator: Identical-looking placebo tablets

Total blood B12 in people with low baseline levels2/4 LowRead moreCollapse
Confidence
2/4 LowOne small study compared two active regimens, with no group that received zero B12. Of the 40 participants, 36 completed testing. A change in a blood marker alone cannot establish whether participants felt any different.

Both regimens raised blood B12 levels, with higher levels on the weekly dose.

Study findings. Mean B12 rose above 240 pmol/L in both groups. From day 30 through day 90, it was higher with 2,000 µg once a week than with 50 µg daily (p<0.01 for between-group comparisons at those time points).

Studied regimen50 µg daily (350 µg per week)Cyanocobalamin, sublingual tablets
Duration and population90 days40 vegans and vegetarians with B12 levels <220 pmol/L. Of these, 36 completed the study, 18 in each group.

Comparator: 2,000 µg of cyanocobalamin once a week and placebo on the other 6 days

Active B12 in the blood (holotranscobalamin)2/4 LowRead moreCollapse
Confidence
2/4 LowOne small study compared two active regimens, with no group that received zero B12. Of the 40 participants, 36 completed testing. A change in a blood marker alone cannot establish whether participants felt any different.

Active B12 in the blood increased with both regimens.

Study findings. Holotranscobalamin, the fraction of B12 bound to a transport protein, increased in both groups (p<0.0001 for change over time). The study found no convincing difference between the regimens.

Studied regimen50 µg daily (350 µg per week)Cyanocobalamin, sublingual tablets
Duration and population90 days40 vegans and vegetarians with B12 levels <220 pmol/L. Of these, 36 completed the study, 18 in each group.

Comparator: 2,000 µg of cyanocobalamin once a week and placebo on the other 6 days

Methylmalonic acid in people with low B122/4 LowRead moreCollapse
Confidence
2/4 LowOne small study compared two active regimens, with no group that received zero B12. Of the 40 participants, 36 completed testing. A change in a blood marker alone cannot establish whether participants felt any different.

The marker of B12 deficiency decreased with both regimens.

Study findings. Methylmalonic acid, a marker that can rise when B12 is low, decreased in both groups (p<0.0001 for change over time). The study found no convincing difference between the regimens.

Studied regimen50 µg daily (350 µg per week)Cyanocobalamin, sublingual tablets
Duration and population90 days40 vegans and vegetarians with B12 levels <220 pmol/L. Of these, 36 completed the study, 18 in each group.

Comparator: 2,000 µg of cyanocobalamin once a week and placebo on the other 6 days

Homocysteine in people with low B122/4 LowRead moreCollapse
Confidence
2/4 LowOne small study compared two active regimens, with no group that received zero B12. Of the 40 participants, 36 completed testing. A change in a blood marker alone cannot establish whether participants felt any different.

Homocysteine decreased with both regimens.

Study findings. Blood homocysteine decreased in both groups (p<0.0001 for change over time). The study found no convincing difference between the regimens. This was a blood test result. The test was not used to assess how participants felt.

Studied regimen50 µg daily (350 µg per week)Cyanocobalamin, sublingual tablets
Duration and population90 days40 vegans and vegetarians with B12 levels <220 pmol/L. Of these, 36 completed the study, 18 in each group.

Comparator: 2,000 µg of cyanocobalamin once a week and placebo on the other 6 days

Pain associated with diabetic nerve damage in people taking metformin2/4 LowRead moreCollapse
Confidence
2/4 LowOne study with 90 participants. It measured many outcomes, and it is unclear whether their order of importance was specified in advance. The study did not report an exact confidence interval for the difference in changes. It is difficult to tell from the study how noticeable the change would be to an individual.

The pain score decreased slightly during usual treatment.

Study findings. On the painDETECT neuropathic pain questionnaire, scores fell from 18.4 to 17.1 points with B12 and rose from 19.3 to 20.9 with placebo. The changes were −1.3 and +1.6 points, respectively (adjusted between-group p<0.001).

Studied regimen1,000 µg per day in addition to usual diabetes treatmentB12fix methylcobalamin, orally dissolving tablets
Duration and population12 months90 people with type 2 diabetes and peripheral and autonomic nerve damage. They had taken metformin for ≥4 years and had B12 levels <400 pmol/L.

Comparator: Placebo tablets in addition to usual diabetes treatment

Blood B12 with oral treatment and injections in people aged 65 and older2/4 LowRead moreCollapse
Confidence
2/4 LowIn the single open-label study, participants knew how they received B12. Of the 283 participants, 229 completed one year of follow-up. The 95% confidence interval for the difference in proportions at one year ranged from −16.6 to +2.9 percentage points. The study compared specific regimens, including infrequent oral dosing after week eight.

At 8 weeks, oral treatment was found to be no worse than injections. At one year, this was not confirmed.

Study findings. By week 8, B12 had reached ≥211 pg/mL in 133 of 140 participants taking oral B12 and 129 of 143 receiving injections. After a year, the figures were 103 of 140 (73.6%) and 115 of 143 (80.4%), respectively. Oral treatment met the prespecified noninferiority criterion at week 8. At one year, it failed to meet that criterion.

Studied regimenBy mouth: 1 mg daily for the first 8 weeks, then 1 mg once a weekOptovite cyanocobalamin, solution from ampoules taken by mouth
Duration and population52 weeks283 people aged 65 and older with B12 levels <211 pg/mL. Of these, 140 received oral B12 and 143 received injections. 229 completed the year of follow-up.

Comparator: The same cyanocobalamin by intramuscular injection, 1 mg every other day in weeks 1–2, once a week in weeks 3–8, and once a month in weeks 9–52

Methylmalonic acid in people with borderline B12 levels3/4 ModerateRead moreCollapse
Confidence
3/4 ModerateThe result concerns a specific blood test. Symptom relief was outside its scope. The study included 50 people, and blood test results at one month were available for 45. The decrease in the marker was confirmed. The study did not demonstrate that the effect persisted long term after discontinuation.

The marker of B12 deficiency decreased during treatment.

Study findings. After one month, methylmalonic acid was 0.23 µmol/L with B12 and 0.37 µmol/L with placebo. The difference adjusted for baseline was −0.13 µmol/L (95% confidence interval −0.19 to −0.06). Three months after treatment ended, the difference was no longer detectable.

Studied regimen1,000 µg per dayB12, oral tablets
Duration and population4 weeks of treatment, with repeat blood testing 3 months after treatment ended50 adults aged 18–91 with B12 levels of 125–200 pmol/L at screening

Comparator: Placebo tablets

Safety and limitations

No tolerable upper intake level has been established for B12 because of its low toxicity. The daily requirements in the table and the high doses used in studies are different reference points. A trial dose of 1,000 µg does not set a standard for everyday intake. If you have a known cobalt allergy, discuss the B12 product with your doctor before taking it. Reactions to cyanocobalamin and hydroxocobalamin can include a rash or hives that appear immediately or 12–72 hours later. If you develop a rash, contact your doctor. Seek urgent help for a widespread blistering rash, difficulty breathing, wheezing, or feeling faint. A cobalt allergy limited to skin reactions does not automatically rule out B12 treatment when it is needed. Numbness or tingling in the arms and legs may occur with B12 deficiency, even without anemia. These symptoms warrant an evaluation to identify the cause. After stomach or intestinal surgery, with impaired absorption, or in older age, an otherwise adequate dietary intake of B12 may be insufficient. Getting enough B12 is especially important during pregnancy and breastfeeding. If a mother following a vegan diet is low in B12, her breastfed child may also be affected. Discuss the B12 source and dose for the mother or child with a healthcare professional. Avoid liver and liver products during pregnancy because they contain high levels of vitamin A.

Vitamin B12 — Health Professional Fact Sheet (NIH ODS) ↗ · Vitamin B12 (hydroxocobalamin, cyanocobalamin): advise patients with known cobalt allergy to be vigilant for sensitivity reactions ↗ · Foods to avoid in pregnancy — NHS ↗

Interactions

WithEffectActionSource
MetforminLong-term use may reduce B12 absorption and lower B12 levels.Discuss monitoring your B12 levels if you take the medication long term. Do not stop taking your prescribed medication on your own.Vitamin B12 — Health Professional Fact Sheet (NIH ODS) ↗
Stomach acid reducers (PPIs and H2 blockers)Omeprazole and cimetidine, for example, can reduce the release of B12 from food and its absorption.Discuss monitoring your B12 levels if you take the medication long term. Do not stop taking your prescribed medication on your own.Vitamin B12 — Health Professional Fact Sheet (NIH ODS) ↗

Food sources

The table lists selected foods from the USDA SR28 database. Values vary by food type and preparation. Weights refer to the edible portion. For fish and chicken breast, the table gives weights before cooking. This is not a suggestion to eat them raw. The serving sizes illustrate the calculation. Together, 200 g of milk, 170 g of plain low-fat yogurt, and 50 g of boiled egg provide about 2.65 µg of B12 and 273 kcal. Unfortified plant foods are not a reliable source of B12. For fortified products, check the amount on the label.

Food composition

FoodGroupAmountServingCaloriesSources
Clams (edible bivalve mollusks)boiled or steamed, without shellsSeafood98.89 µg / 100 g85 g · 84.1 µg · 126 kcal148 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
Beef liverpan-friedOrgan meats83.13 µg / 100 g50 g · 41.6 µg · 88 kcal175 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
Musselsboiled or steamed, without shellsSeafood24 µg / 100 g85 g · 20.4 µg · 146 kcal172 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
Pork liverbraisedOrgan meats18.67 µg / 100 g50 g · 9.34 µg · 83 kcal165 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
Sockeye salmonedible portion before cookingFish4.69 µg / 100 g150 g · 7.04 µg · 197 kcal131 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
Pink salmonedible portion before cookingFish4.15 µg / 100 g150 g · 6.23 µg · 191 kcal127 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
Wild Atlantic salmonedible portion before cookingFish3.18 µg / 100 g150 g · 4.77 µg · 213 kcal142 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
Egghard-boiledEggs and dairy1.11 µg / 100 g50 g · 0.555 µg · 78 kcal155 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
CheddarcheeseEggs and dairy1.1 µg / 100 g30 g · 0.33 µg · 121 kcal404 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
Plain nonfat yogurtwithout additivesEggs and dairy0.61 µg / 100 g170 g · 1.04 µg · 95 kcal56 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
2% milkfluid, with vitamins A and DEggs and dairy0.53 µg / 100 g200 g · 1.06 µg · 100 kcal50 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗
Chicken breastraw, skinless and bonelessMeat0.21 µg / 100 g150 g · 0.315 µg · 180 kcal120 kcal / 100 gUSDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗

USDA National Nutrient Database for Standard Reference, Release 28 (SR28): ASCII data and documentation ↗

Studies

PublicationWhat was studiedParticipantsForm and doseResult
Surplus vitamin B12 use does not reduce fatigue in patients with Irritable Bowel Syndrome or inflammatory bowel disease: A randomized double-blind placebo-controlled trial ↗Fatigue in people with bowel conditions95 adults aged 18–65 with marked fatigue and either irritable bowel syndrome or inflammatory bowel disease. Baseline B12 was ≥150 pmol/L.Oral B12, 1,000 µg per day, for 8 weeks. Comparator: Placebo.Fatigue decreased by 8.1 points with B12 and 8.3 points with placebo on the CIS subjective fatigue subscale. B12 provided no additional reduction in fatigue.
Effects of vitamin B-12 supplementation on neurologic and cognitive function in older people: a randomized controlled trial ↗Memory in people aged 75 and older with moderately low B12201 people aged 75 and older with B12 levels of 107–<210 pmol/L at screening, without anemia or marked memory impairment. Cognitive data were available for 184 participants.Cyanocobalamin oral tablets, 1 mg (1,000 µg) per day, for 12 months. Comparator: Identical-looking placebo tablets.On a word-list memory test, the B12 group scored an average of 1.4 words lower than the placebo group after adjustment for baseline scores. The 95% confidence interval ranged from −2.9 to +0.1 words. The study found no advantage for B12 over placebo in memory performance or on the other cognitive tests studied.