What is the best way to measure vitamin B12?
Measuring serum B12 alone misses a true deficiency too often; when in doubt, combine it with MMA and take kidney function and any supplementation into account.
Measuring serum B12 alone gives an incomplete picture. Roughly half of the people with a subclinical deficiency still get a 'normal' result. You can therefore miss a deficiency if you look at this number alone.
Two additional markers are more sensitive: methylmalonic acid (MMA) and homocysteine. Both rise early when the body cannot use enough B12 properly, even before serum B12 itself becomes abnormal. In daily practice, the combination of total serum B12 and MMA is the most useful: a low B12 provides an initial indication, and a high MMA confirms the deficiency. Bear in mind the person's age and kidney function, because impaired kidney function also raises MMA independently of B12.
Another option is holotranscobalamin, the fraction of B12 that is actually available to cells. A study of 231 healthy older adults found that holotranscobalamin correlated more strongly with measurable neurological changes than total serum B12. That said, these additional markers are not perfect either: they produce inconsistent results, and diagnosing a B12 deficiency remains difficult even when multiple markers are combined.
One noteworthy side point: an unexpectedly high B12 value is usually harmless, but can sometimes point to an underlying cancer. In such cases the high figure is caused by more carrier proteins in the blood, not by more available B12. A specialised laboratory technique that was previously used to investigate this further produces misleadingly inaccurate results too often and is no longer recommended.
Are you already taking high-dose B12 supplements? Then a blood test at that moment has little value: the result will be distorted by what you have recently taken. Stopping supplementation and then re-testing gives a more reliable picture. Finally, the same study in older adults suggests that even values within the current 'normal' range can be associated with neurological changes, both at the low and the high end. Whether the normal threshold in use warrants adjustment remains a matter of debate.
Based on two clinical review studies and one observational study in 231 healthy older adults. No RCTs are available; all evidence is associative in nature.