B12 and the nervous system: the absorption chain, and who actually runs short

The treatment arrived twenty-two years before the molecule

Vitamin B12 has one of the longest absorption chains of any nutrient, and every link in it can fail. This is what the vitamin does for nerve tissue, who genuinely runs short, and the limit that most B12 marketing steps over: fixing a deficiency and improving on adequacy are not the same experiment.

A disease that was fatal, and then was not

Pernicious anaemia killed people. It progressed over years, produced anaemia and neurological decline, and nothing reliably stopped it.

In 1926 George Minot and William Murphy reported that feeding patients large daily quantities of raw or lightly cooked liver reversed it (Journal of the American Medical Association, 1926; 45 patients). They shared the 1934 Nobel Prize in Physiology or Medicine with George Whipple for the work.

The active compound was not isolated until 1948. For twenty-two years the treatment was half a pound of liver a day, and it worked.

The treatment arrived twenty-two years before the molecule
Figure 1 The treatment arrived twenty-two years before the molecule Minot and Murphy reversed pernicious anaemia with a daily liver diet in 1926. Vitamin B12 itself was not isolated until 1948. Minot and Murphy, Journal of the American Medical Association, 1926 · 45 patients.

Two reactions, and only two

In human biochemistry B12 is a cofactor for exactly two enzymes. Everything attributed to the vitamin runs through one of them.

Methionine synthase

This converts homocysteine to methionine, which becomes S-adenosylmethionine — the body's general-purpose methyl donor. Without B12 the reaction stalls, homocysteine accumulates, and methylation reactions across the body slow down.

Methylmalonyl-CoA mutase

This handles a step in the breakdown of certain fatty acids and amino acids. When it stalls, methylmalonic acid accumulates — which is why methylmalonic acid is a more specific marker of B12 status than a blood B12 level.

Two enzymes, two accumulating markers
Figure 2 Two enzymes, two accumulating markers B12 serves methionine synthase and methylmalonyl-CoA mutase. When each stalls, homocysteine and methylmalonic acid respectively build up — which is how deficiency is confirmed when a blood level is ambiguous. A diagram of established biochemistry.

Why nerve tissue is where it shows first

Myelin, the insulating sheath around nerve fibres, is lipid-rich and methylation-dependent to build and maintain. A methylation bottleneck reaches it early.

Prolonged deficiency produces subacute combined degeneration of the spinal cord — demyelination of the dorsal and lateral columns, which carry position sense and motor signals. The clinical picture is numbness, unsteadiness and weakness.

The finding that changed how it is diagnosed

For decades the assumption was that anaemia comes first and neurology follows, so a normal blood count ruled it out.

In a series of 141 consecutive patients with confirmed B12 deficiency and neuropsychiatric problems, 40 of them — 28% — had neither anaemia nor enlarged red cells (Lindenbaum et al., New England Journal of Medicine, 1988).

That result is why B12 deficiency is now investigated on the neurological picture rather than waiting for the blood count to move.

Twenty-eight per cent had a normal blood count
Figure 3 Twenty-eight per cent had a normal blood count Of 141 consecutive patients with confirmed B12 deficiency and neuropsychiatric findings, 40 had neither anaemia nor macrocytosis. Lindenbaum et al., New England Journal of Medicine, 1988 · n=141 · consecutive clinical series.

The absorption chain is unreasonably long

The stomach has to release it from food

B12 in food arrives bound to protein. Stomach acid and pepsin have to free it before anything else can happen — the link that fails first when the chain fails, and the commonest failure in older adults.

The stomach also has to make intrinsic factor

Parietal cells in the stomach lining secrete intrinsic factor, a protein that binds B12 and escorts it onward. Pernicious anaemia is the autoimmune destruction of those cells.

The complex has to survive the journey

Bound to intrinsic factor, B12 then travels the length of the small intestine, and the complex has to arrive at the far end intact.

The terminal ileum has to take it up

Absorption happens at one specific site near the end of the small intestine, through a dedicated receptor. Disease or surgery there removes the route entirely.

Four links, four ways to fail
Figure 4 Four links, four ways to fail Release from food protein, binding to intrinsic factor, transit, then uptake at the terminal ileum. A supplement bypasses the first step and nothing else. A diagram of established physiology.

The chain, and what breaks each link

Step What has to happen What interferes
Release from food Acid and pepsin free B12 from dietary protein Atrophic gastritis, acid-suppressing medication
Intrinsic factor Gastric parietal cells secrete the carrier protein Autoimmune destruction, gastric surgery
Transit The bound complex travels the length of the small intestine to the terminal ileum Intestinal disease or surgery along the way
Ileal uptake A receptor at the terminal ileum takes up the complex Crohn's disease, ileal resection
Passive diffusion About 1% of a dose crosses without any of the above Nothing — it just needs a large enough dose

Why age is the dominant risk factor

Atrophic gastritis reduces acid and pepsin, which leaves B12 stuck to food protein. Between 10% and 30% of older adults may malabsorb food-bound B12 for this reason (NIH Office of Dietary Supplements, citing the clinical literature).

The consequence is specific and slightly odd: those people absorb B12 from a supplement or fortified food well, because it is not protein-bound — though the rest of the chain still has to work. The problem is the meat, not the tablet.

Metformin, measured in a trial rather than assumed

Metformin interferes with B12 absorption, and unusually for this subject the effect has been quantified in a randomised trial rather than inferred.

Over 4.3 years of metformin against placebo in people with type 2 diabetes, mean B12 concentration fell by about 19% relative to placebo, and the absolute risk of B12 deficiency rose by roughly 7 percentage points (de Jager et al., BMJ, 2010; n=390, randomised placebo-controlled).

A 19% fall, over four years, in a randomised trial
Figure 5 A 19% fall, over four years, in a randomised trial Metformin against placebo lowered mean vitamin B12 concentration by about 19% and raised the absolute risk of deficiency by about 7 percentage points. de Jager et al., BMJ, 2010 · n=390 · type 2 diabetes · 4.3 years.

Who is genuinely at risk

Older adults with reduced stomach acid. People on long-term metformin or acid-suppressing medication. Anyone who has had gastric or ileal surgery.

And people eating no animal products, because B12 in the food supply is bacterial in origin and reaches humans through animal tissue. A systematic review of vegetarian and vegan populations reported deficiency rates ranging from around a tenth to the large majority of a group depending on age and diet strictness (Pawlak et al., Nutrition Reviews, 2013; review of published prevalence studies).

The width of that range is itself the finding. It reflects genuinely different diets and different cut-offs, and any single headline percentage drawn from it is being quoted out of a distribution.

The risk factors, medical and dietary
Figure 6 The risk factors, medical and dietary Long-term metformin, acid-suppressing medication, gastric or ileal surgery and atrophic gastritis are the common medical causes of shortfall; a strict plant-based diet is the main dietary one. Underlying sources named in the surrounding text.

The one per cent that makes oral doses work

Alongside the intrinsic-factor route there is a second, long-established one: roughly 1% of an oral dose crosses the gut wall by passive diffusion, needing no acid, no carrier protein and no receptor — a figure established in early absorption work and carried forward in standard nutrition references.

That inefficiency turns out to be useful. In patients with confirmed deficiency, oral cobalamin at 2,000 mcg a day was at least as effective as intramuscular injection over four months (Kuzminski et al., Blood, 1998; n=38, randomised).

One per cent of 2,000 mcg is 20 mcg, which comfortably exceeds a daily requirement. The route is bad and the dose makes up for it.

Why 500 mcg prints as a five-figure percentage

The label Daily Value for B12 is 2.4 mcg. Divide 500 by 2.4 and the panel prints 20,833%.

That number is arithmetic against a small reference figure, not a measure of potency — the same trap as any other percentage column, unpacked in how to read a Supplement Facts panel in about sixty seconds. Amounts well above the requirement are conventional here precisely because absorption is unreliable, which is a reason and not a boast.

A large percentage against a very small denominator
Figure 7 A large percentage against a very small denominator The label Daily Value for B12 is 2.4 mcg, so 500 mcg prints as 20,833%. The size of the figure reflects the size of the reference amount. Declared per 21 CFR 101.36. A panel in the format the rules require.

The limit

Correcting a deficiency works

That has been established since 1926 and is not seriously contested. Neurological recovery is often partial when treatment starts late, which argues for finding it early rather than for taking more.

Adding B12 to someone already replete is a different experiment

Healthy older adults with raised homocysteine were given folate, B12 and B6 for two years. Homocysteine fell as intended. Cognitive test performance did not differ from placebo (McMahon et al., New England Journal of Medicine, 2006; n=276, aged 65 and over).

The concession is real: a trial in people with mild cognitive impairment and raised homocysteine reported a slower rate of brain atrophy on the same class of intervention (Smith et al., PLOS ONE, 2010; n=168, two years). That population had a diagnosis, and the result does not transfer to people who do not.

Homocysteine fell, cognition did not move
Figure 8 Homocysteine fell, cognition did not move Two years of folate, B12 and B6 in healthy adults aged 65 and over with raised homocysteine lowered homocysteine without improving cognitive test performance against placebo. McMahon et al., New England Journal of Medicine, 2006 · n=276 · two years.

The form in the brackets

B12 appears on panels as methylcobalamin, cyanocobalamin, adenosylcobalamin or hydroxocobalamin. Cyanocobalamin is the most-studied and the most stable; methylcobalamin is one of the two forms the body actually uses.

The comparison is narrower than either camp claims, and it is set out with the evidence in methylcobalamin versus cyanocobalamin. What matters more on a label is that the form is printed at all, rather than buried — the problem described in what a proprietary blend hides.

The only vitamin that contains a metal
Figure 9 The only vitamin that contains a metal B12 is built around a cobalt ion, which is where the deep red colour of the isolated vitamin comes from and why the family is named cobalamin.

Where we stand

Weal Focus contains 500 mcg of B12 as methylcobalamin, printed with the form named and outside any blend. It is there because intake and absorption vary from person to person and 500 mcg covers the gap at negligible cost, not because more of it makes anyone sharper.

Someone with symptoms suggesting a deficiency needs a blood test and a clinician, not a supplement bought on the strength of an article. The everyday pattern of losing words mid-sentence has many causes and this is only one of them. The same insurance-not-enhancement reasoning applied to a mineral is in why a zinc shortfall is more common than it sounds, and the question of how long any of this takes to register is in how long a supplement actually takes.

These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.

Sources for the claims and figures on this page, with the population studied in each, are on our citations page.