Vitamin B12 for Neuropathy
Vitamin B12 reliably improves nerve symptoms only when blood levels are deficient or borderline; for people with normal B12, the benefit for general or diabetic neuropathy is modest and far from proven. The evidence is preliminary, so get tested before supplementing.
Why this grade7-layer evidence engine
The grade is B (preliminary) because the trials lean positive but the benefit is conditional on B12 status. The strongest signal is in diabetic peripheral neuropathy: a meta-analysis of 14 RCTs (PMID 37189436, n=1303) found methylcobalamin sped nerve conduction (motor +3.69 m/s, sensory +3.96 m/s, all p<0.001) and improved symptoms (RR 1.27). A 12-month placebo-controlled RCT in metformin-treated, B12-deficient diabetics (PMID 33042414, n=90) likewise improved sensory conduction, pain, and quality of life (p=0.001).
The benefit is largest when a deficiency is actually corrected. A deficiency-neuropathy review (PMID 38034928) reported 70-85% of confirmed-deficient patients improved within 3-6 months, with better recovery if treated early. For pain specifically, a systematic review (PMID 33285134, n=768) found consistent relief in diabetic and idiopathic cases but weaker, variable effects in post-herpetic and chemotherapy-induced neuropathy. An earlier review (PMID 32281030, n=372) saw symptom gains in 3 of 4 trials but inconsistent nerve-conduction results.
Regulators and clinics draw the same line. The NIH Office of Dietary Supplements states B12 has 'no beneficial effect... in the absence of a nutritional deficit,' and the American Academy of Neurology and ADA (L5e) recommend B12 only when lab-confirmed deficiency is the cause, not as a general neuropathy treatment. Mayo and Cleveland Clinic warn that untreated deficiency causes nerve damage but do not endorse B12 for non-deficient patients. Practical takeaway: test serum B12 (and MMA) first; if you are not deficient, B12 is unlikely to fix the problem and is not a substitute for first-line care.
Scoring transparency
All scores computed by a 7-layer evidence engine — fully auditable▸View the full decision path (audit trail)
- compute_raw_score — 加權公式: L2×0.30 + L3×0.25 + L5×0.25 + L11×0.10 + L1×0.10 = 0.558
- tier_from_score — 依分數區間映射至 tier letter
- apply_hec_rules — 高品質 SR/MA 顯示 positive (3 篇 > 0 negative)
- tier_strict_requirement_check — Tier 條件達標,未降階
- detect_disputes — 偵測到 0 個 hard + 0 個 soft dispute
- decide_status — 依 tier + dispute 結果決定 status