Vitamin B12: What the Tests Actually Measure

The clinical question with vitamin B12 is rarely "does this person need more?" but "does this person have enough?" Those are different questions, and they are answered by different tests. A brief on what each test actually measures, where the reference ranges mislead, and what repletion trials show about how to top someone up.

What the standard tests measure

Total serum B12 is the default screen. It captures cobalamin bound to two transport proteins — haptocorrin, which carries roughly 80% of circulating B12, and transcobalamin, which is the fraction actually delivered to cells. Most B12 in serum sits on haptocorrin and is metabolically inactive. So a "normal" total can sit on top of cellular insufficiency, and a "low" total can still reflect adequate cellular delivery when the transcobalamin-bound fraction is preserved.

Methylmalonic acid (MMA) and homocysteine are the functional markers. Both rise when B12-dependent enzymes lose activity at the cellular level. Elevated MMA is the more specific signal — homocysteine also climbs in folate deficiency, kidney impairment, and several genetic variants. A guideline-compliant workup for ambiguous cases pairs serum B12 with MMA, sometimes with holotranscobalamin (the "active" B12 fraction) as a third axis.

Where reference ranges mislead

The lab reference range for serum B12 is usually around 200–900 pg/mL. The lower bound is a population threshold, not a physiological one — it describes roughly the bottom 2.5% of a screened population, not a point where biology changes. Studies of older adults have repeatedly reported neurological and hematologic findings in people with serum B12 in the 200–350 range, inside the reference window. Functional markers sometimes flag deficiency at concentrations a routine panel would call normal.

The implication is not that everyone with a low-normal B12 needs treatment. It is that the reference range answers "where do you sit relative to the screened population?" — not "is your B12 status sufficient for your tissues?"

Who actually gets deficient

Three populations show up repeatedly in the literature.

• Older adults, often with atrophic gastritis. Stomach acid is required to release dietary B12 from food protein; reduced acid production cuts absorption from food but generally not from supplements, where the B12 is already free.
• Long-term metformin users, where the absorption interference appears dose- and duration-dependent.
• Long-term proton pump inhibitor users, again via reduced gastric acid.

Strict vegan diets without supplementation also produce deficiency over years, but the pattern there is straightforward dietary insufficiency rather than absorption-limited.

What repletion trials actually show

The most cited comparison is oral versus intramuscular cobalamin for established deficiency. A Cochrane review pooled the small randomized trials available and concluded that high-dose oral cobalamin (1000–2000 mcg/day) normalized serum B12 to a degree comparable with IM injections at 1 and 4 months. Subsequent observational and trial data have largely supported this, with two caveats: oral repletion may take longer to resolve neurological symptoms in severe deficiency, and adherence is the dominant variable when comparing routes in the real world.

The form question — cyanocobalamin vs. methylcobalamin vs. hydroxocobalamin — generates more marketing copy than evidence. All three normalize standard markers when dosed adequately. Specific scenarios, like certain inborn errors of cobalamin metabolism, do favor non-cyano forms, but those are exceptions rather than the general consumer case.

Energy, mood, and cognition claims

A common reason readers ask about B12 is fatigue or brain fog. The trial evidence here is unsympathetic: in adults without measurable deficiency, B12 supplementation does not improve energy, mood, or cognitive performance in randomized trials. The exception is the deficient subgroup, where repletion does help — but the relevant question becomes "are you actually deficient?" rather than "does B12 help?"

A reading checklist

When evaluating a new B12 paper or product claim:

• Did the study measure functional markers (MMA, homocysteine), or only total serum B12?
• Was the reference range population-derived or symptom-anchored?
• Were participants drawn from a known at-risk group, or the general population?
• Was supplementation oral, IM, or sublingual — and was adherence tracked?
• Were neurological outcomes assessed, or only hematologic ones?
• Did the trial enrich for deficient participants, or test repletion in already-replete subjects?

Most B12 confusion in popular health writing comes from skipping one of those distinctions. The trials are mostly clear; the translation isn't.