The Vitamin B12 Test Your Doctor Orders May Be Measuring the Wrong Thing for Your Brain

Your doctor told you your B12 is fine. Your brain may not agree.

When 231 healthy older adults walked into a UCSF lab for a brain aging study, their B12 levels looked unremarkable by every standard measure. The average was 414.8 picomoles per liter of blood, far above the U.S. deficiency threshold of 148 (Annals of Neurology). By every normal reference range, they were not just adequate — they were comfortably supplied.

Then researchers measured the form of B12 that actually gets into the brain. The picture changed. Participants with lower levels of that active fraction, called holo-transcobalamin, showed slower processing speed on cognitive tests, delayed responses to visual stimuli, and a higher volume of white matter lesions on MRI scans (Annals of Neurology). The effect was stronger with older age.

The paper, published in February 2025 in the Annals of Neurology, did not prove that low active B12 causes cognitive decline. What it showed is that your doctor can order a standard blood panel, get a reassuring number, and have no idea that your neurons may be running short.

Here is the problem: the B12 test ordered in every routine blood panel measures total serum cobalamin — all the B12 in your blood, regardless of whether it is bound to the protein that carries it into the central nervous system. The biologically active form, holo-transcobalamin, is what your brain actually uses. A result in the normal range tells you that you have not hit deficiency. It does not tell you that your brain is getting what it needs.

Labcorp and Quest Diagnostics both offer holo-transcobalamin testing. It is not a secret assay. But it is not what your doctor receives on a standard metabolic panel. Quest's own internal guidance, reviewed by this newsroom, lists total serum B12 as the appropriate test for routine screening and reserves holoTC for specific follow-up scenarios (Quest Diagnostics). If your number comes back above 200 picograms per milliliter, the reflex to active B12 typically does not fire. That is where millions of older Americans may be sitting in a diagnostic blind spot.

The 148 picomole per liter cutoff used in the United States was never calibrated for brain health outcomes. It was set to catch frank deficiency — the level below which megaloblastic anemia and irreversible neurological damage appear. Researchers have argued for years that the threshold for neurological risk sits higher, and that the population of people between not deficient and optimally supplied may be larger than clinicians appreciate. The UCSF data is consistent with that view: the participants with the worse cognitive and MRI profiles were not deficient by any current standard. They were simply lower within the normal range.

The comparison that specialists reach for is vitamin D. For decades, 20 nanograms per milliliter was treated as sufficient. Then research linked lower levels to bone density loss, autoimmune disease, and cardiovascular risk, and clinical reference ranges crept upward. The process was slow, contested, and never fully resolved. B12 may be on a similar trajectory, and the lab testing infrastructure has not caught up with the science.

There are important caveats. The study is observational and cannot establish causation. The cohort was 231 people, a sample size that supports the signal but does not close the question. A 2025 Mendelian randomization study found no protective effect of higher total serum B12 on cognitive outcomes — though that analysis used total B12, not the active fraction, which is precisely the distinction the UCSF team is arguing matters (SciTechDaily). And even if active B12 does turn out to be the relevant biomarker, supplementation trials in people who are not clearly deficient have not yet shown clear cognitive benefit. The biological mechanism is plausible. The clinical evidence is still building.

What is clear is that the test your last physical ordered is not telling you what the latest neuroscience suggests it should. Ari Green, the senior author and a neurologist at UCSF's Weill Institute for Neurosciences, put it plainly in the university's press release: healthy B12 levels may not be enough to ward off neurocognitive decline (UCSF News). That is a different bar than the one your lab report is built around.

The solution, if one exists, is not a mystery. Better testing is available. What is missing is the guideline infrastructure to make it standard, and the clinical awareness to order it before symptoms appear. Until that changes, a normal result on a routine panel is not the clean bill of brain health it appears to be.