High-dose B vitamin supplementation and cognitive decline in Alzheimer disease: a randomized controlled trial.

Randomized, placebo-controlled, multicenter trial conducted at 40 U.S. sites in adults with probable mild to moderate Alzheimer disease. Eligibility included age over 50 years and MMSE 14 to 26; the active treatment arm included 240 randomized participants, with 204 completing the final study visit and 218 analyzed for outcomes.

The active regimen was oral high-dose B vitamin supplementation for 18 months: folic acid 5 mg/day, vitamin B12 1 mg/day as cyanocobalamin, and vitamin B6 25 mg/day as pyridoxine hydrochloride. Participants in the active treatment arm were randomized against placebo in a 3:2 allocation.

High-dose B vitamins lowered homocysteine but did not slow cognitive decline or improve clinical outcomes over 18 months. Change in ADAS-cog was not different between active treatment and placebo (7.38 [9.72] vs 6.54 [8.17]; P = .52), and the primary adjusted analysis was also nonsignificant (P = .56). Other outcomes were similarly null, including MMSE change (-2.15 [4.21] vs -2.51 [3.59]; P = .69), CDR sum of boxes change (2.58 [2.45] vs 2.51 [2.57]; P = .57), NPI change (3.31 [12.84] vs 2.20 [11.12]; P = .28), and ADCS-ADL change (-10.96 [12.36] vs -10.00 [11.09]; P = .42). Time-to-endpoint analysis was also null (hazard ratio 0.99; 95% CI 0.80 to 1.21; P = .97), while depression adverse events were more common with active treatment (67 [27.9%] vs 30 [17.8%]; P = .02).

The trial found no clinical benefit despite biomarker improvement, and the 18-month duration may have been insufficient to detect a slower neurodegenerative effect. Attrition and analysis counts were incomplete relative to randomization, and the depression adverse-event imbalance raises safety concerns. Generalizability may be limited to older adults with mild to moderate Alzheimer disease in a folate-fortified U.S. setting.