Vitamin B12 after fifty: why absorption drops and what the trials used

Age rarely raises how much B12 you need. What it breaks is the step that pries B12 loose from the protein in food. Among adults over 60, those with a worn-out stomach lining showed signs of deficiency 38% of the time, against 15% of everyone else.¹ The B12 in tablets and fortified cereal skips that step entirely.¹

The bottleneck is the stomach, not the plate

B12 in meat, fish, eggs and dairy arrives stuck to protein. Stomach acid and pepsin, the enzyme that cuts protein apart, have to free it first. Only then can intrinsic factor — a carrier molecule the stomach also makes — pick it up and hand it to the gut.

When acid production falls, that first step fails and everything after it keeps working. The condition behind most of this is atrophic gastritis: the stomach lining thins with age and stops producing acid. It turned up in 15% of 3,299 adults over 60.¹

A study from 1991 shows the split cleanly. Eight older adults with atrophic gastritis and eight normal controls swallowed labeled B12 in both forms.² The atrophic gastritis group absorbed 0.7% of the protein-bound dose, against 1.9% in the controls. The crystalline form was absorbed just as well by both groups.²

That single experiment is the practical point of the whole topic. A person who cannot get B12 out of a boiled egg can still absorb it from a fortified cereal or a tablet, because the crystalline form was never bound to protein to begin with.² A review of how the body handles B12 with age puts it bluntly: too little in the diet has not proved to be a major cause, and malabsorption is what produces deficiency in older people.⁸

Acid-blocking drugs reproduce the same failure

The clearest confirmation comes from people whose stomach acid was suppressed on purpose. In older people taking omeprazole, absorption of protein-bound B12 dropped to 0.50%, against 1.21% in people not taking it.³

Two details make the mechanism hard to argue with. Taking the dose with cranberry juice raised absorption. Dilute hydrochloric acid raised it further.³ Put the acid back and the vitamin comes free.

The same pattern shows in the large group of older adults. Proton pump inhibitors had been taken for at least six months by 37% of them, and were linked to lower levels of the usable fraction of B12 in blood — though only among those on 30 mg a day or more.¹ Deficiency ran at 38% in the atrophic gastritis group, 21% among the drug users and 15% in everyone else.¹

A 2025 review gathers the causes into one list: gastric changes with age, acid suppression by drugs, metformin, and atrophic gastritis that is not autoimmune.⁴ All of them differ from pernicious anemia, which destroys intrinsic factor rather than acid.

Fortified food helps, and not enough for the worst affected

If crystalline B12 sidesteps the problem, fortified food ought to protect people, and it partly does. Eating it regularly — five or more portions a week rather than four or fewer — went with better B12 readings across the whole group.¹

In the people who actually had atrophic gastritis, it was not enough to bring status back to normal. The authors' conclusion was that foods for this group may need more added B12 than they currently carry.¹

What the dosing evidence actually consists of

Less than you would expect for a vitamin this well known. A review of the trials found eight that qualified, and opened by explaining why the recommendations rest on so little: most studies gave B12 by injection, or at doses several hundred times the recommended intake.⁶

The blood markers themselves work. Serum B12, methylmalonic acid and homocysteine all responded when intake changed in people with low or borderline status.⁶ What the reviewers asked for was trials of small oral doses across the whole age range — the doses ordinary food and fortified cereal actually deliver.

Where the evidence is better is the route. Three trials in 153 people, treated and followed for three to four months, compared swallowing B12 with injecting it.⁵ At 1000 μg a day — 1 mg, the amount in a high-dose tablet — there was no clinically meaningful difference in blood levels between the two.⁵ One trial at 2000 μg, twice that, favored the tablet by 680 pg/mL.⁵

The reviewers graded that evidence low quality, because the numbers were imprecise. They noted the tablets cost less. And they pointed out that not one trial reported whether the symptoms of deficiency actually went away.⁵

The amounts used for the age-related version of the problem are much smaller than replacement injections. A daily 125 to 250 mcg by mouth has shown efficacy as an alternative to injections.⁴ A review of screening in older adults suggested 1 mg daily for a month, then 125 to 250 µg as maintenance for dietary shortfall, and 1 mg daily for pernicious anemia.⁷ These are treatment regimens for a diagnosed deficiency, not intakes for people whose readings are fine.

Diagnosis is the weak link

The reason this subject stays muddled is that nobody agrees where the line sits. There is no consensus or guideline for diagnosing B12 deficiency.⁷ Setting a blood cut-off is difficult. Homocysteine and methylmalonic acid pick up problems earlier, but they can return false results, and their normal ranges are not standardized.⁷

Deficiency can exist while blood B12 sits above the traditional lower limit, and the normal ranges for both of those markers shift with age and sex.⁸

Which is why the same review that lists every argument for screening older adults — more of them affected, symptoms easy to miss, treatment cheap and safe — still records that no mass screening is recommended.⁷ The test is not good enough to screen with.

What remains unknown

The unglamorous questions are the open ones. Small oral doses have never been tested across the age range, which is exactly the range fortified food operates in.⁶ How much extra B12 a fortified food would need to carry to fix status in atrophic gastritis has not been worked out.¹

The 1991 study also found that antibiotics restored protein-bound absorption, which points at gut bacteria as part of the problem.² Nobody developed that into a treatment. Nothing here compares the two common tablet forms, cyanocobalamin and methylcobalamin.

And while older patients recover from deficiency as completely as younger ones, in both blood counts and nerve function, long-standing dementia responds poorly.⁸ That sets a limit on what correcting a number can undo.