Prospective study of plasma folate, vitamin B12, and cognitive function and decline.

BACKGROUND: The relation between B vitamins and cognitive decline is controversial. In this study, we explored the association of plasma folate and vitamin B12 with cognitive function measured approximately 10 years later. METHODS: We determined plasma folate and vitamin B12 levels from blood samples collected in 1989 to 1990 and initially evaluated cognition in 1995 to 2001 among 635 women, age 70+ years, from the Nurses' Health Study. In a subset of 391, 3 repeated cognitive tests were completed for evaluation of cognitive decline over 4 years; repeated testing is ongoing for the remaining women. Our primary outcome was a global composite score of 6 neuropsychologic tests administered by telephone. We used linear regression models to estimate multivariable-adjusted mean cognitive performance across quartiles of the vitamins and longitudinal models for cognitive decline. RESULTS: Higher vitamin levels were not associated with either initial cognitive performance or subsequent cognitive decline. Mean difference in initial global score for top versus bottom quartiles was 0.06 standard units for folate (95% confidence interval [CI] = -0.10 to 0.22) and 0.15 units for vitamin B12 (0.00 to 0.31). There were no dose-response trends for either nutrient. Women with high levels of both nutrients initially performed better than women low in both nutrients (global score, mean difference = 0.34; 95% CI = 0.05 to 0.62); this association did not hold for subsequent cognitive decline. CONCLUSIONS: Combined B vitamin deficiency may be associated with impaired cognition, but in these healthy, well-nourished women, plasma folate and vitamin B12 were not related to cognitive function.