B12: 1000 mcg vs 2000 mcg, and Whether Double Helps

Not in this trial. Both doses improved the numerical rating scale for pain and the Michigan neuropathy screening instrument examination score with statistical significance, and the paper reports no significant between-group differences. That is the central finding and it is easy to miss, because both arms improving looks like a success for both doses when what it actually shows is that the extra 1000 mcg added nothing measurable over 16 weeks.

Three things. The numerical rating scale for pain fell from 7.00 to 5.60 in the 1000 mcg group (P = 0.016) and from 6.18 to 4.42 in the 2000 mcg group (P = 0.007). The Michigan neuropathy screening instrument examination score improved from 5.70 to 5.22 at 1000 mcg (P = 0.033) and from 5.40 to 4.47 at 2000 mcg (P = 0.022). The neuropathy disability score did not change in either group, at P greater than 0.05.

People with diabetic peripheral neuropathy and low serum vitamin B12, which is a specific and important qualification. Thirty-five were randomly assigned and 32 completed the 16-week trial, with 1:1 allocation. Every conclusion here applies to that population: adults with an existing deficiency and an existing nerve complication. It says nothing about what happens when someone with normal B12 takes a supplement.

No, and this is the most important boundary on the page. The participants were selected for low serum B12. Correcting a deficiency and topping up an adequate level are different interventions with different expected results, and research on the first is routinely quoted as though it supported the second. If your B12 is in range, this trial does not tell you that supplementing will do anything at all.

The neuropathy disability score did not change in either group over 16 weeks. That is worth taking seriously rather than explaining away. It suggests the improvement seen was in symptoms and in one screening examination rather than in the broader disability measure, and 16 weeks may simply be too short for structural nerve change to register. A trial that improves some outcomes and not others is more informative than one that improves everything, because it shows the measures were capable of not moving.

That is a question for whoever ordered your blood test, because the answer depends on why your level is low. Absorption is the usual issue: B12 requires intrinsic factor and a functioning terminal ileum, and people who cannot absorb it well are exactly the people who end up on high oral doses or injections. This trial suggests 1000 mcg performed as well as 2000 mcg in its population; it does not establish a dose for anyone outside it, and this article is not medical advice.

B12 is water-soluble and generally regarded as having low toxicity, with excess excreted rather than stored to harmful levels, which is why very high doses are sold without much concern. That is a reasonable safety picture, but it is not a reason to take more. The finding here is about futility rather than danger: if the doubled dose produced no measurable benefit over 16 weeks, the extra is being paid for and passed through.

Modest weight, carefully placed. It is a randomised trial with a clear comparison and pre-specified measures, which is a good design, and it found a null result between groups. Small trials are more likely to miss a real difference than to invent one, so a null finding in 32 people leaves open the possibility that a larger study would find a small advantage for the higher dose. What it does not support is the current retail assumption that the higher dose is obviously better.