Did you know…

The standard blood test for vitamin B12 deficiency is a late, relatively insensitive, and unspecific biomarker. In studies comparing it against clinical outcome, its sensitivity has ranged from 13% to 75%. Meaning: in the worst studies, seven out of eight B12-deficient patients had a normal serum B12. The functional markers that detect deficiency earlier and more reliably exist, are validated, and are almost never ordered in first-line clinical workup.

The claim, in the reference of laboratory medicine

A 2020 diagnostic-accuracy analysis published in an international lab-medicine journal, drawing on more than 11,000 samples, reported the definitive comparison:

“Total serum Vitamin B12 is a late, relatively insensitive and unspecific biomarker of deficiency.”

The area-under-the-curve comparison in that same analysis was:

• Holotranscobalamin (holoTC, active B12): 0.92
• Methylmalonic acid (MMA): 0.91
• Total serum B12: 0.90
• Homocysteine: 0.78

Why the earlier markers are not ordered

Total serum B12 measures all forms of circulating cobalamin, including the roughly 80% that is bound to haptocorrin and is not bioavailable to cells. Only the fraction bound to transcobalamin (holoTC) can be taken up and used. Methylmalonic acid rises as the enzymatic step that requires B12 in the mitochondrion begins to stall. Both markers detect functional deficiency before circulating B12 drops out of the reference range. Both are validated. Both are almost never ordered in first-line clinical workup, because serum B12 is cheaper, faster, and highly automated. The clinical guidance in most jurisdictions defaults to the cheaper marker, and the deficit that presents late becomes the clinical picture that gets diagnosed late.

The neurological cost of the false reassurance

B12 deficiency can produce cognitive impairment, peripheral neuropathy, and demyelination of the spinal cord before it produces the megaloblastic anaemia that older textbooks treat as the diagnostic signature. In a population masked by folate fortification (which normalizes the anaemia without touching the neurological lesion) and reassured by a normal serum B12 (which the assay is too insensitive to catch early), the neurological damage accumulates while every routine number reads normal.

This entry is a companion to arguments on reference ranges calibrated on populations rather than on function and to The Hunger We Don’t See.