Summarized & reviewed by The Peptide Dispatch Editorial Team · Last reviewed September 14, 2026
Vitamin B12 is one of the few micronutrients that shows up on a routine panel without anyone asking for it. It gets ordered for fatigue, for numbness, for a high mean cell volume, or as part of an annual workup, and the result comes back with a reference range that runs from roughly 200 to 900 picograms per milliliter depending on the laboratory. A value of 260 is reported as normal. A value of…
This dispatch covers Your B12 Is Normal. The Floor of That Range Misses Nerve Damage, the Assay Can Be Fooled, and the Test That Settles It Is Almost Never Ordered. in the research research category, authored by The Peptide Dispatch Editorial Team. Estimated reading time: 11 minutes. The Peptide Dispatch curates peer-reviewed peptide research for self-directed learners. All summaries are presented for Research Use Only and do not constitute medical advice.
Vitamin B12 is one of the few micronutrients that shows up on a routine panel without anyone asking for it. It gets ordered for fatigue, for numbness, for a high mean cell volume, or as part of an annual workup, and the result comes back with a reference range that runs from roughly 200 to 900 picograms per milliliter depending on the laboratory. A value of 260 is reported as normal. A value of 310 is reported as normal. Nobody looks at it again.
That habit rests on three assumptions, and each of them has been contradicted in the published literature for decades. The first is that a serum B12 result above the laboratory floor rules out deficiency. The second is that the assay itself reliably measures what it claims to measure. The third is that if B12 were really a problem, the blood count would show it. This article takes the three in order.
Serum B12 is a total. Most of the cobalamin circulating in blood is bound to a carrier protein called haptocorrin, and cobalamin bound to haptocorrin cannot be taken up by most tissues. Only the fraction bound to a second carrier, transcobalamin, is available to cells. That fraction, called holotranscobalamin or active B12, is typically somewhere between a fifth and a third of the total the laboratory reports.
The practical consequence is that two people with an identical serum B12 of 300 can have very different amounts of usable cobalamin. Conditions that raise haptocorrin, including liver disease and certain blood disorders, can push the total up while the active fraction stays flat or falls. The reverse happens too. A serum total sitting in the lower part of the normal range says very little on its own about whether the cell has enough.
This is why the functional markers exist. When cells run short of usable cobalamin, two reactions that depend on it slow down. One converts methylmalonyl-CoA to succinyl-CoA, and when it stalls, methylmalonic acid (MMA) accumulates and spills into blood and urine. The other converts homocysteine to methionine, and when it stalls, homocysteine rises. MMA is the more specific of the two, because homocysteine also climbs when folate or B6 is short. MMA rises for essentially one nutritional reason, and it rises before the blood count changes.
The clearest demonstration of the problem with the 200 floor is now nearly forty years old. In 1988 a group at Columbia-Presbyterian published in the New England Journal of Medicine a series of 141 consecutive patients whose neurological or psychiatric abnormalities were caused by cobalamin deficiency. Forty of them, 28 percent, had no anemia and no macrocytosis. The hematocrit was normal in 34, the mean cell volume was normal in 25, and both were normal in 19. Their presentations included paresthesia, sensory loss, ataxia, dementia, and psychiatric disorders, some longstanding.
Among those 40 patients, the serum cobalamin was below 100 pg/mL in only 22. Sixteen sat between 100 and 200, and two were above 200, inside what most laboratories then and now would report as normal. What the whole group had in common was markedly elevated methylmalonic acid and homocysteine. Every patient who survived the first week of treatment improved neurologically, and MMA or homocysteine fell by more than half in every patient measured.
Read that carefully. More than a quarter of patients with documented neurological damage from B12 deficiency had a normal blood count, and nearly half of that subgroup had a serum B12 that a routine panel would not flag or would flag only as borderline. The functional marker, not the serum total, was the finding that made the diagnosis.
Later work established the gray zone this creates. Serum values from roughly 200 up to somewhere between 300 and 400 pg/mL are the band where a meaningful fraction of people have elevated MMA and are functionally deficient, and where the serum result alone cannot tell you which ones. Some guideline bodies now describe this band explicitly as indeterminate and recommend MMA to resolve it. Most laboratory reports still print a single lower limit and a single word next to the value.
The second assumption, that the number is at least an honest measurement of total B12, failed in a way that reached the New England Journal of Medicine in 2012.
The immunoassays most laboratories now use to measure B12 are competitive-binding assays that rely on intrinsic factor as the binding protein. Pernicious anemia, the classic cause of severe B12 deficiency, is an autoimmune condition in which the patient produces antibodies against intrinsic factor. Those antibodies do not disappear when the serum is put into an analyzer. Carmel and Agrawal reported that in a subset of patients with confirmed pernicious anemia, the automated assays returned normal or falsely elevated B12 results because the patient's own antibodies interfered with the assay's binding step.
The interference is not exotic. It is the disease's defining feature colliding with the assay's design. A patient with the single most important cause of severe deficiency can receive a result that says nothing is wrong. Laboratories and manufacturers have worked on this since, and the problem is better recognized than it was, but the underlying mechanism has not gone away, and the result that comes back on a routine panel does not carry a warning label.
For anyone with a clinical picture that fits deficiency and a serum B12 that does not, the appropriate response is not to trust the number. It is to measure MMA, and where pernicious anemia is a real possibility, to test for intrinsic factor antibodies directly rather than relying on a B12 assay that those antibodies may be corrupting.
The third assumption is the one that keeps the first two from being noticed. Clinicians were trained that B12 deficiency produces macrocytic anemia, so a normal hemoglobin and a normal mean cell volume are taken as reassurance.
The 1988 series already showed that 28 percent of patients with neurological damage had neither. Later reviews from the same group put the general figure for neurological involvement without hematologic change at a substantial minority of all cobalamin-deficient patients, and noted an inverse relationship that is still not fully explained: the patients with the least anemia often had the most neurological damage, and the severity of neurological abnormalities before treatment correlated with how long the symptoms had been present.
Two other things blunt the blood count as a signal. Folic acid, which is added to grain products in many countries and present in most multivitamins, corrects the anemia of B12 deficiency without correcting the neurological process. And iron deficiency, which is common, pushes the mean cell volume down while B12 deficiency pushes it up, so the two can cancel and leave a normal-looking average.
The upshot is that a normal complete blood count does not exclude B12 deficiency, and waiting for it to become abnormal means waiting past the point at which the neurological findings are most reversible.
Two very widely prescribed drug classes have population-level data behind them.
In the Diabetes Prevention Program Outcomes Study, published in the Journal of Clinical Endocrinology and Metabolism in 2016, participants randomized to metformin at 850 mg twice daily were followed for up to 13 years. At five years, low B12 (at or below 203 pg/mL) was found in 4.3 percent of the metformin group against 2.3 percent on placebo. When the borderline band was included (at or below 298 pg/mL), the figures were 19.1 percent against 9.5 percent at five years and 20.3 percent against 15.6 percent at thirteen. Each year of metformin use was associated with a 13 percent increase in the odds of deficiency. Peripheral neuropathy was more prevalent in metformin users with low B12. The authors' conclusion was that routine B12 testing in metformin-treated patients should be considered. Note the numbers: one in five long-term metformin users was low or borderline, and the borderline band is exactly where the serum test cannot tell you whether the cell is short.
Acid suppression is the other. A Kaiser Permanente case-control study published in JAMA in 2013 compared 25,956 patients with an incident diagnosis of B12 deficiency against 184,199 without. A two-year or longer supply of a proton pump inhibitor carried an odds ratio of 1.65 for deficiency, and a two-year supply of an H2 blocker carried 1.25. Higher PPI doses, above 1.5 pills a day, carried an odds ratio of 1.95. The mechanism is not mysterious. Dietary B12 is bound to protein, and releasing it requires stomach acid. Suppress the acid and the vitamin passes through.
Beyond those two, the list of people who plausibly sit in the gray zone is long: adults over 60, in whom atrophic gastritis and food-cobalamin malabsorption become common; anyone eating little or no animal protein; anyone who has had bariatric surgery or extensive small-bowel disease; anyone whose appetite or intake has dropped sharply for any reason, including pharmacological appetite suppression; and anyone exposed to nitrous oxide, which chemically inactivates cobalamin and can produce classic neuropathy with a normal or near-normal serum level. Several of these overlap in the same person.
Given all of this, here is a reasonable way to interpret a serum B12 on a standard panel.
A result well above 400 pg/mL, in someone without a clinical picture suggesting deficiency and without a reason for assay interference, is reassuring. A result below 200 is deficiency until proven otherwise and does not need a second marker to justify attention. The band in between is the problem, and in that band the serum value should be treated as a screening result, not an answer.
Methylmalonic acid is the test that resolves it. An elevated MMA with a low-normal B12 indicates that the cell is short regardless of what the total says. A normal MMA with a low-normal B12 is a much more reassuring pair. MMA has one important confounder: it rises in reduced kidney function independently of B12 status, so it should be read alongside creatinine or cystatin C. Homocysteine adds information but is not specific, and holotranscobalamin, where available, measures the usable fraction directly and can be a useful second line.
If the picture is deficiency and the B12 result is normal, the assay may be the problem, and intrinsic factor antibody testing bypasses it.
None of this requires a new panel. B12 is already there. The reference range printed beside it was designed around the anemia that deficiency causes late, not the neurological damage it causes early, and the number sits on a measurement that counts cobalamin your cells cannot reach and can be defeated by the antibodies that define its most important cause. A follow-up test that costs little and has existed since the 1980s answers the question the serum total cannot. It is almost never ordered, and the reason is that the word next to the number says normal.
Based on articles retrieved from PubMed:
Educational content only. This article describes published research and laboratory interpretation. It is not medical advice and does not recommend any test, supplement, or treatment for any individual. Discuss your own results with a qualified clinician.
Educational content — not medical advice. Effects described are drawn from cited research in study subjects.