B vitamins and homocysteine: what changed in humans?
In this investigation
The question for this food log
I have no recorded homocysteine measurement or diagnosed B-vitamin deficiency. My food estimates cannot supply either. The intake references for folate, B12 and B6 remain separate from the question of taking extra vitamins to improve health or training.
A positive brain scan is a specific result
VITACOG randomized 271 adults aged at least 70 with mild cognitive impairment to folic acid, B12 and B6 or placebo for two years. Its primary MRI analysis used 168 participants with usable paired scans: annual brain shrinkage averaged 0.76% versus 1.08%. This was a positive imaging result, not a dementia-prevention endpoint. The related cognitive report used the same trial; it is not independent replication. The study disclosed pharmaceutical support and patent interests. Its subgroup findings also need the paper’s multiple-comparison caveat. Original MRI trial
Lower homocysteine did not consistently improve cognition
B-PROOF’s author flow chart shows 3,027 older adults with elevated homocysteine assigned to folic acid and B12 or placebo, with 2,919 starting treatment. Both groups received vitamin D for two years. Homocysteine fell more with the B vitamins, but four cognitive domains did not improve. Cognitive outcomes were secondary to the fracture question. Only usable follow-up tests entered these analyses; 2,556 supplied the global screening score. Its roughly 0.2-point advantage would disappear after correction for multiple tests, the authors noted. Public and commercial funding were disclosed. Original cognitive report, author manuscript and flow
A newer positive trial still has a boundary
A 2025 trial randomized 220 adults aged 40–80 with cerebral small-vessel disease and cognitive impairment to 0.8 mg folic acid daily or placebo for six months. The primary MoCA screening score favored folic acid by 1.27 points (95% CI 0.52–2.03). Trail-making tests, mobility and quality of life did not significantly improve. Follow-up was incomplete, the report’s completion counts are inconsistent, and its sample-size calculation used a different cognitive test. Registry history was unavailable for this review. This is an encouraging result in a clinical population, not demonstrated dementia prevention or improved cognition in healthy young adults. Original trial
Stroke prevention benefited one defined population
CSPPT randomized 20,702 Chinese adults with hypertension and no previous stroke or heart attack. Adding 0.8 mg folic acid to enalapril, compared with enalapril alone, reduced first strokes over a median 4.5 years: 2.7% versus 3.4% (hazard ratio 0.79; 95% CI 0.68–0.93). This was the primary clinical outcome; blood pressure was similar between groups. The trial described a setting without widespread folic-acid fortification. It stopped early under its reported monitoring rules; funders, including the product company, participated in its design. These circumstances matter when transferring the result to another population. Original trial
Kidney disease gave a warning against marker-only reasoning
DIVINe randomized 252 people with diabetic kidney disease to a high-dose folic-acid/B6/B12 combination or placebo; 238 entered the modified intention-to-treat analysis. Homocysteine fell, yet the primary measured kidney-filtration outcome worsened more with treatment: the between-group decline differed by 5.8 mL/min/1.73 m² at 36 months (95% CI 1.1–10.6). Missing measurements required replacement and imputation. A secondary vascular/death composite also favored placebo, with an imprecise nominal result. The study disclosed donated tablets and author financial interests. This does not show that ordinary food folate harms kidneys; it shows why a favorable marker cannot stand in for the clinical outcome. Original trial
What changes here
These trials establish neither a B-vitamin stack nor a homocysteine target for this routine. Their positive findings deserve their population and endpoint; their null findings do not prove equivalence. A diagnosed deficiency would be a different clinical question.