B Vitamins

Most healthy adults eating a mixed diet don't need a B-complex pill — but four groups need to watch their B12, the B6 dose on the label matters more than the form, and the cognitive payoff in trials is very small.

The eight B vitamins are coenzymes: small molecules that enzymes need in order to run the reactions of energy metabolism, DNA synthesis, and the clearing of a toxic amino acid called homocysteine. Classic deficiency — pellagra, beriberi, the anemia of B12 depletion — is rare in well-fed populations, so the modern question is not "am I deficient?" but "do I belong to a group whose intake or absorption is quietly compromised, and if so, what should I actually take?" The honest answer is that a generalised "B-complex for longevity" is mostly wasted money, while a few targeted uses are genuinely worth the trouble.

Who actually needs them

Strong that four groups need to watch their B12: vegans should supplement, the others should get checked or take a low dose; for everyone else, supplementing delivers no measurable benefit.

For a healthy adult eating animal products and a reasonable variety of plants, the B vitamins are abundant and the body recycles them efficiently. Four situations change that:

  1. Vegetarians and especially vegans. Plant foods contain essentially no vitamin B12 — it is made by bacteria and reaches us through animal products. This group should supplement B12 without waiting for a test.
  2. Long-term metformin users. Metformin interferes with the calcium-dependent step that lets the gut absorb the B12–intrinsic-factor complex, and in the large US diabetes-prevention cohort, followed for about 13 years, each extra year of metformin use was linked to about 13% higher odds of low B12, and low or borderline-low B12 was more common in the metformin group (20% against 16%).[1]
  3. People on chronic acid suppression. Proton-pump inhibitors and H2-blockers cut the stomach acid needed to free B12 from food protein. People taking both metformin and a proton-pump inhibitor have two separate reasons to get B12 checked; an analysis of US drug-safety reports hinted the combination may be worse, but found no clear evidence that the two drugs interact.[2]
  4. Adults past about 60. Age-related atrophic gastritis becomes more common, impairing the acid and intrinsic factor that dietary B12 needs; it affects about 8–9% of adults over 65, against 2% of the general population.[3]

Outside these groups, mega-dosing a B-complex does not measurably improve energy, mood, or cognition in people who aren't deficient — a point worth holding onto, because the category is marketed almost entirely on the opposite promise.

Cognition: a very small effect, and one brain-scan trial

Strong evidence that the overall benefit is very small; larger effects only in subgroups such as older adults with mild cognitive impairment and elevated homocysteine.

The headline reason people take B vitamins for the aging brain is homocysteine. This amino acid, a by-product of methionine metabolism, is toxic to neurons and blood vessels at elevated levels and is an independent risk factor for brain shrinkage and cognitive decline. Vitamins B6, B9 (folate), and B12 are the obligatory cofactors that clear it from the blood, so lowering homocysteine with B vitamins is biologically plausible.

The evidence is more conditional than the marketing. The largest recent meta-analysis, 17 trials in about 5,300 adults aged 60 or older, found a very small benefit of B6, folate or B12 on overall thinking ability once outlier and single-blind trials were set aside, probably too small to notice in daily life, and similar whether people had normal cognition, mild impairment or dementia.[4] The meta-analysis did not test whether people with high homocysteine gain more; the strongest subgroup signal comes from the brain-scan trial below, which measured brain shrinkage rather than thinking ability.

The single trial that anchors the optimistic case is VITACOG, a British study in adults with mild cognitive impairment in which high-dose B6/B9/B12 slowed the whole-brain atrophy rate by 29.6% (0.76%/yr active vs 1.08%/yr placebo, P=0.001), rising to a 53.3% reduction in those in the top homocysteine quartile (>13 µmol/L).[5] A later re-analysis sharpened the catch: the atrophy benefit appeared only in participants who already had high blood levels of omega-3. In those with good omega-3 status, B vitamins slowed brain shrinkage by about 40% (and up to ~46% in the subgroup with the most EPA, eicosapentaenoic acid, in their blood); in those with low omega-3 it did essentially nothing.[6] The companion analysis found the same omega-3 dependence for cognitive decline.[7] These figures come from a trial powered for atrophy rather than cognition, with unadjusted subgroup p-values, so the precise magnitudes are fragile, and the omega-3 dependence has not been confirmed in other trials (see below). Later metabolomic work on the same trial showed the B vitamins were reshaping homocysteine and neurotransmitter-related pathways, not acting through a single mechanism.[8]

The practical message: the B-vitamin cognitive intervention is worth considering mainly in older adults with documented high homocysteine. In VITACOG the brain-shrinkage benefit appeared only in people with good omega-3 levels, but post-hoc checks of two other trials in middle-aged and older adults with raised homocysteine did not confirm this cleanly: in one (folic acid alone) people with low omega-3 benefited more, and in the other only DHA showed any link.[9][10] It remains an unconfirmed subgroup finding, so pairing B vitamins with adequate omega-3 is reasonable rather than required. See Omega-3 and Dementia prevention.

Vitamin B12: testing and form

Strong for repletion in the at-risk groups; the case for premium forms is Weak.

In recent US national survey data, where fortified foods and supplements are common, about 4% of adults, and about 4% of over-60s, had clearly low blood B12, and about 1 in 8 were borderline.[11] Blood B12 can miss a tissue shortfall (see below), and the reason to watch B12 after 60 is that absorbing it from food gets harder, not a fixed deficiency rate. One under-appreciated interaction sharpens the masking concern: when B12 status is low, a high serum folate has been associated with worse cognition and higher homocysteine and methylmalonic acid — a reason the fortification era makes adequate B12 matter more, not less.[12]

Serum B12 is an insensitive test. The body defends blood levels at the expense of tissue stores, so a "normal" serum B12 can sit on top of genuine intracellular depletion. The functional markers — methylmalonic acid (MMA) and homocysteine — rise before serum B12 falls, which is why ambiguous cases should be confirmed with MMA or holotranscobalamin rather than serum B12 alone. In a US survey sample of 747 adults with heart failure, those with the highest MMA had about 50% higher risk of death than those with the lowest (hazard ratio 1.52), even though blood B12 itself was not linked to death.[13] MMA also rises with poor kidney function and with age, so a high value needs reading alongside kidney function.[14] The testing cadence for at-risk groups lives in Midlife labs.

On the form, the marketing runs ahead of the evidence. B12 is sold cheaply as cyanocobalamin and at a premium as methylcobalamin (the active coenzyme form). There is no evidence that the body absorbs supplemental B12 differently by form, or that sublingual tablets beat swallowed ones — cheap oral cyanocobalamin works fine for most people.[15] Methylcobalamin's theoretical advantage (it skips a conversion step and acts as a direct methyl donor) is real biochemistry but has not translated into a meaningful clinical edge for routine repletion. If you want it, the cost is modest; just don't expect the form alone to do anything dramatic.

Dose: 25–250 µg/day for prevention in an at-risk group; 1000 µg/day for documented deficiency. High oral doses work even when absorption is impaired by drugs or atrophic gastritis, because 1–2% crosses the gut by passive diffusion independent of intrinsic factor. B12 is water-soluble with no established upper limit — oversupplementation is not a meaningful toxicity concern.

Folate: methylfolate versus folic acid

Weak for choosing 5-MTHF over folic acid (either form works); Strong for folic acid in neural-tube-defect prevention.

Natural food folate, the synthetic folic acid in fortified foods and cheap supplements, and 5-methyltetrahydrofolate (5-MTHF, the active circulating form) are not chemically identical, though for most people taking a supplement they do the same job.

Folic acid is biologically inert until the body reduces it, a multi-step conversion whose first enzyme (dihydrofolate reductase) has low and variable capacity in humans. High intakes can outrun it, leaving unmetabolised folic acid circulating in the blood — a state of uncertain long-term significance that can also mask the anemia which normally signals a hidden B12 deficiency, delaying its diagnosis. A large share of people carry a common variant in the MTHFR gene that slows the conversion of folate to its active form; in the United States more people carry one or two copies of the C677T variant than don't. But the CDC states that people with these variants can still process all types of folate, including folic acid, and that the variant is not a reason to avoid it, and a US medical genetics society advises against MTHFR testing as part of a routine work-up for clotting risk.[16][17] In a randomised trial in 144 healthy women, 5-MTHF lowered homocysteine about as well as folic acid.[18] Its real advantages are that it needs no conversion step and leaves no unmetabolised folic acid in the blood: in a small single-dose study of one 5-MTHF salt in 24 healthy adults, funded by the salt's maker, it was absorbed better than folic acid and unmetabolised folic acid did not rise.[19] No trial shows better health outcomes with 5-MTHF, so either form is fine when folate is actually needed; 5-MTHF is a reasonable choice, not a necessity.

For pregnancy the answer is settled: women who could become pregnant should ensure adequate folate for neural-tube-defect prevention, and folic acid is the form with the deepest evidence base in that specific context. The case for it is overwhelming at the population scale — mandatory folic-acid fortification cut neural-tube defects by 46% in Canada (1.58 to 0.86 per 1,000 births) and by roughly a quarter to a half across other fortifying countries.[20]

On the cancer question that the folic-acid debate inevitably raises: a randomised polyp-prevention trial (n=1,021) found that 1 mg/day folic acid did not prevent colorectal adenomas and was associated with more advanced and multiple lesions.[21] A pooled analysis of 13 randomised trials (about 50,000 people, mostly in heart-disease prevention trials) found that folic acid supplements, at doses roughly ten times what fortification provides, made no significant difference to colorectal or overall cancer rates over about five years of treatment.[22] The defensible reading is that high doses warrant some caution, but no colorectal-cancer effect has been shown in trials.

Vitamin B6: the one with a real toxicity ceiling

Caution — documented harm (peripheral neuropathy) above the intake limit; the cancer and fracture signals are Weak/observational.

B6 is the B vitamin most worth being careful with, because more is not safer. Chronically exceeding the intake limit over months to years can cause a sensory peripheral neuropathy — tingling and numbness in the hands and feet. Symptoms usually stop if the supplement is discontinued as soon as they appear[23], but recovery is not guaranteed: in Australia's 2025 review of vitamin B6, 103 people reported severe and sometimes permanent effects, including nerve damage and muscle weakness.[24] The European Food Safety Authority sets the tolerable upper limit for adults at 12 mg/day, counting food and supplements together, far below the doses packed into many "energy" and "stress" B-complex formulas.[25] Australia, whose own upper level is 50 mg/day, has required a neuropathy warning on products above 10 mg/day since 2022,[26] and in November 2025 decided that, from 1 June 2027, products with more than 50 mg per daily dose will need a pharmacist's advice to buy (above 200 mg they already need a prescription). Its regulator concluded that neuropathy can occur at 50 mg/day or even lower, and Australia's own 50 mg/day limit is itself under review, due by early 2027.[27] If you take a B-complex, check the B6 content against the 12 mg/day EFSA ceiling. The active form, pyridoxal-5-phosphate (P5P), is sometimes preferred but does not exempt you from the dose caution.

Two observational signals reinforce why stand-alone high-dose B6/B12 products deserve particular suspicion — both are association data, vulnerable to confounding, not proof of harm. In the VITAL cohort (n=77,118), long-term individual-supplement use among men was tied to roughly doubled lung-cancer risk (B6 >20 mg/day HR 1.82, 95% CI 1.25–2.65; B12 >55 µg/day HR 1.98, 95% CI 1.32–2.97) — the 95% confidence interval is the range where the true effect most plausibly lies, and because both intervals sit entirely above 1.0, chance is an unlikely explanation — highest in current smokers, with no association in women and none from multivitamins.[28] Later work pooling blood samples from 20 cohorts, plus a genetic (Mendelian randomisation) analysis, found that people with higher blood B12 levels had somewhat more lung cancer, though the genetic result was borderline.[29] These studies measured blood levels, not supplement doses, so they lend some support to caution about high-dose B12 without proving harm. And in the Nurses' Health Study (n=75,864), combined high intake of B6 and B12 was associated with increased hip-fracture risk — though B12 alone carried no such signal, so older adults who genuinely need B12 should not be deterred.[30] The recurring theme: caution is warranted when supplementing without a documented deficiency.

Niacin, NAD+, and the longevity-molecule question

Weak for the NAD+-booster longevity claims; a Caution signal now attaches to niacin/NAD+ metabolites.

Vitamin B3 is the precursor for NAD+ (nicotinamide adenine dinucleotide), the coenzyme central to energy metabolism, DNA repair, and the sirtuin enzymes that feature heavily in aging research. This is the biochemistry behind the NMN (nicotinamide mononucleotide) and nicotinamide riboside (NR) supplements sold as anti-aging compounds — and it is where the gap between mechanism and proven human benefit is widest. These precursors reliably raise blood NAD+, but the functional results in people have been small and inconsistent, and the longevity claims rest on animal data and surrogate markers rather than hard outcomes. The full, skeptical treatment of NMN and NR is in Geroprotectors; the short version is that the hype substantially outruns the evidence.

The older, cheaper end of B3 has a longer track record — and it has not aged well. Plain niacin (nicotinic acid) raises NAD+ severalfold and once anchored lipid therapy, but the two large modern trials retired it: AIM-HIGH was halted early for futility in 2011, and HPS2-THRIVE (n=25,673) found niacin added no cardiovascular benefit on top of statins while raising new-onset diabetes, infection, and bleeding.[31] Its use is also limited by an uncomfortable prostaglandin-mediated flush, and its EFSA upper limit is correspondingly low at 10 mg/day for supplemental nicotinic acid. The other B3 form, nicotinamide, carries a much higher limit (900 mg/day). Neither is a longevity intervention for a healthy adult on current evidence. Nicotinamide does have one niche with trial support: in people who had already had two or more non-melanoma skin cancers, nicotinamide 500 mg twice daily cut new ones by about a quarter over a year in a randomised trial.[32] That is a treatment for people with a skin-cancer history, not a longevity supplement; the trials, including a null one in transplant recipients, are covered in Nicotinamide.

A 2024 finding complicates the whole NAD+-booster story. The terminal niacin/NAD+ catabolites 2PY and 4PY were associated with elevated three-year major-adverse-cardiovascular-event risk across a discovery and two validation cohorts, and 4PY drove vascular inflammation in mice through an endothelial adhesion molecule (VCAM-1) that recruits inflammatory cells to the artery wall — a mechanistic plus associational signal, not yet interventional evidence.[33] Ferrell's paper notes that NR and NMN also raise these metabolites, so the finding is a reason for caution about the NAD+ boosters, not proof of harm. A 2025 health-records study of 13,108 patients (91% men, average age 67), about 5,300 of whom took nicotinamide, found no overall increase in major cardiovascular events (heart attack, stroke or cardiovascular death), though some estimates were imprecise.[34]

A further note: the long-held belief that NAD+ inevitably falls with age has been challenged. A 2026 study of seven groups of people found that NAD+ in whole blood stays stable with age and does not shift with lifestyle interventions, though nicotinamide riboside supplements do raise it.[35] Blood is not muscle or brain, and levels may still fall in some tissues and diseases, but this undercuts the simple "top up a universal deficit" rationale for NAD+ boosters in the well.

The rest of the complex

Weak as a longevity case, with two well-evidenced niche uses: riboflavin for migraine prophylaxis (Moderate) and thiamine repletion in defined at-risk groups (Strong).

The remaining B vitamins are essential coenzymes but rarely the limiting factor in a mixed diet, and none has earned a stand-alone longevity case:

  • B1 (thiamine) feeds carbon into the energy-producing TCA (tricarboxylic acid) cycle and is critical for nerve function; deficiency is reversible and frequently missed. The at-risk set is broader than alcohol alone: chronic alcohol use disorder (the classic cause of Wernicke encephalopathy), pre- and post-bariatric surgery, refeeding syndrome, and heart failure on chronic loop diuretics — though the prevalence in stable outpatient heart failure is genuinely uncertain, with modern estimates well below the high figures from older literature.
  • B2 (riboflavin) is a core electron carrier in the mitochondrial respiratory chain. In lab-grown human cells, extra riboflavin in the culture medium made cells less likely to become senescent after treatment with a DNA-damaging drug; this has not been tested in animals or people.[36] It may also lower blood pressure in people with two copies of the MTHFR 677TT variant (about 10% of the population): in one trial of 91 such people already treated for high blood pressure, a small daily dose of riboflavin (1.6 mg) for 16 weeks lowered systolic pressure by about 6 mmHg more than placebo.[37] It is an early example of nutrition targeted to genotype rather than a general recommendation. Separately, riboflavin has a well-established and cheap use it is rarely credited for: at 400 mg/day it roughly quadrupled migraine responders versus placebo (59% vs 15%, number-needed-to-treat 2.3).[38]
  • B5 (pantothenic acid) is the backbone of coenzyme A, needed for fat metabolism; dietary deficiency is almost unheard of.
  • B7 (biotin) supports gluconeogenesis and participates in DNA-repair-related histone modification. Supplemental megadoses, common in hair-and-nail products, are mostly harmless but interfere with many laboratory immunoassays — including thyroid and troponin tests — and should be stopped before bloodwork.

EFSA has not set an upper limit for B1, B2, B5 or B7 because there is not enough data to derive one, which is not the same as proof they are harmless at any dose. For B12 it found no defined adverse effects.

Cardiovascular: prevention, not reversal

Strong for no effect on heart attacks or death; a small stroke benefit (about 10%), seen mainly where food is not fortified with folic acid.

Higher dietary intake and blood levels of several B vitamins (B1, B2, B3, B6, B9) correlate with a 10–20% lower stroke risk across two large US cohorts — but this is an observational dietary association, not evidence that taking the pills lowers risk.[39] The randomised evidence is the more important and more sobering fact: the large trials that successfully lowered homocysteine with B vitamins did not reduce heart attacks or cardiovascular death. HOPE-2 (n=5,522) cut homocysteine but left the primary cardiovascular composite unchanged;[40] NORVIT and SEARCH in post-MI patients and VITATOPS in prior-stroke patients were likewise null. A 2017 Cochrane review of 15 trials (about 71,000 adults) found no effect on heart attacks or death, but a small reduction in stroke: 4.3% of people on B vitamins had a stroke against 5.1% on the comparison, about 10% lower in relative terms (risk ratio 0.90).[41] Homocysteine looks more like a marker of vascular risk than a lever you can pull to change outcomes, and trials aimed at acutely reversing established arterial stiffness have been mixed at best.[42]

The one apparent exception sharpens the rule. CSPPT randomised 20,702 Chinese hypertensive adults — a population with low baseline folate and no food fortification — to enalapril with or without folic acid, and the folic-acid arm cut first stroke by 21% (HR 0.79; 2.7% vs 3.4%) over 4.5 years, concentrated in those with the lowest folate.[43] That is a deficiency-correction effect in an unfortified population, not a generalisable benefit for already-replete Western adults — which is exactly why the null Western trials and this positive one are consistent. Pooled across 21 trials (about 116,000 people), folic acid lowered stroke risk by about 10%; the benefit was seen in regions without, or with only partial, grain fortification (about 17% lower risk), while in regions with full fortification there was no reduction.[44] B vitamins belong in the prevention-of-deficiency column, not the cardiovascular-drug column.

Practical summary

  • Don't take a generic high-dose B-complex for longevity. In a non-deficient adult it does nothing measurable, and the B6 load in many formulas is a genuine downside.
  • Vegans and vegetarians: supplement B12 without waiting for a test. Long-term metformin users: diabetes guidelines advise considering periodic B12 checks, especially with anaemia or nerve symptoms.[45] Long-term acid-suppressant users: get checked if you have symptoms such as tiredness, anaemia, or tingling or numbness;[46] the same approach makes sense for people over ~60, though NICE does not list age alone as a risk factor. For any of these groups a low-dose supplement is a cheap, low-risk option. 25–250 µg/day for prevention, 1000 µg/day for documented deficiency.
  • Confirm B12 status with MMA or holotranscobalamin, not serum B12 alone. Homocysteine is a useful readout of B6/B9/B12 status: a raised value is a reason to look for and correct low B12, folate or B6, but no trial shows that pushing an already-normal value lower improves health. See Midlife labs.
  • If you supplement folate, either form works. 5-MTHF is a reasonable choice; folic acid remains the standard for prenatal guidance. Both count toward the same 1,000 µg/day upper limit.
  • The B-vitamin cognitive intervention (B6/B9/B12 in older adults with high homocysteine) is the one cognitive use with trial support; pairing it with adequate omega-3 is reasonable, though the omega-3 link is unconfirmed.
  • Timing doesn't matter much. No trial has compared morning and evening dosing; take them at whatever time you will remember every day.
  • Skip the NAD+ boosters (NMN, NR) for now — see Geroprotectors.
VitaminSensible adult useUpper limit (EFSA, adults)
B6 (pyridoxine / P5P)Only as part of a complex; watch the dose12 mg/day — neuropathy above it
B9 (folate)Either form if supplementing (5-MTHF optional); folic acid 400 µg for possible pregnancy1,000 µg/day (folic acid and 5-MTHF from supplements and fortified foods combined; food folate not counted)
B12 (cobalamin)25–250 µg/day at-risk; 1000 µg/day for deficiencynone established
B3 (nicotinic acid)Not a longevity supplement10 mg/day (flush)
B3 (nicotinamide)—900 mg/day
B1, B2, B5, B7Diet is almost always sufficientnone established

Further reading

  • Smith AD et al. Homocysteine-lowering by B vitamins slows the rate of accelerated brain atrophy in mild cognitive impairment (VITACOG). PLoS ONE 2010.[47]
  • Jernerén F et al. Brain atrophy in cognitively impaired elderly: the importance of long-chain ω-3 fatty acids and B vitamin status in a randomized controlled trial. Am J Clin Nutr 2015.[48]
  • Martí-Carvajal AJ et al. Homocysteine-lowering interventions for preventing cardiovascular events. Cochrane Database Syst Rev 2017 (15 RCTs, 71,422 participants).[49]
  • Ferrell M, Hazen SL et al. A terminal metabolite of niacin promotes vascular inflammation and contributes to cardiovascular disease risk. Nat Med 2024.[50]
  • Brasky TM, White E, Chen CL. Long-Term, Supplemental, One-Carbon Metabolism-Related Vitamin B Use in Relation to Lung Cancer Risk in the Vitamins and Lifestyle (VITAL) Cohort. J Clin Oncol 2017.[51]
  • Aroda VR et al. Long-term Metformin Use and Vitamin B12 Deficiency in the DPP/DPPOS. J Clin Endocrinol Metab 2016.[52]
  • Lamers Y et al. Supplementation with [6S]-5-methyltetrahydrofolate or folic acid equally reduces plasma total homocysteine concentrations in healthy women. Am J Clin Nutr 2004.[53]
  • Berg J et al. Efficacy of B Vitamin Supplementation on Global Cognitive Function in Older Adults: A Systematic Review and Meta-analysis. Nutr Rev 2025.[54]
  • EFSA. Overview on Tolerable Upper Intake Levels. EFSA, version 11, August 2025.[55]
  • Yu W et al. Exploring the relationship between vitamin B12, methylmalonic acid levels and all-cause mortality in heart failure populations: insights from the NHANES database. Front Nutr 2025.[56]
  • Zhang X et al. Intake of B vitamins and their circulating levels in relation to incident stroke in women and men: findings from two national prospective cohorts. Am J Prev Cardiol 2026.[57]

— § —