Heterogeneity and lack of good quality studies limit association between folate, vitamins B-6 and B-12, and cognitive function.

We conducted a systematic review to evaluate the association between folate, vitamin B-6, vitamin B-12, and cognitive function in the elderly. Our search was conducted in Medline for English-language publications of human subjects from 1966 through November 2006; we supplemented these results with information from article reviews and domain experts. We included longitudinal cohort and case-control studies of B vitamins and analyses of cognitive tests or Alzheimer's disease. We evaluated the quality and heterogeneity of study outcomes and assessed 30 different cognitive function tests. Of 24 studies that met eligibility criteria, 16 were determined to be of fair quality. A majority of the studies reviewed 2 or more B vitamins. Considerable heterogeneity was found among B-vitamin-level thresholds, comparisons, and data analyses. Six of 10 folate studies reported a significant association between low baseline blood folate concentrations and subsequent poor test performance in the global cognitive domain, and 4 of 9 folate studies found associations between low blood folate concentrations and increased prevalence of Alzheimer's disease. Studies did not reveal an association of vitamin B-6 and vitamin B-12 blood concentrations with cognitive-test performance or Alzheimer's disease, nor was B-vitamin dietary intake associated with cognitive function. Higher plasma homocysteine concentrations were associated with poorer cognitive function. Although the majority of studies indicated that low blood folate concentrations predicted poorer cognitive function, data supporting this association were limited because of the heterogeneity in cognition-assessment methodology, and scarcity of good quality studies and standardized threshold levels for categorizing low B-vitamin status.