Plasma total homocysteine and memory in the elderly: the Hordaland Homocysteine Study.

We examined the relation between plasma total homocysteine (tHcy), folate, vitamin B12, and episodic memory in elderly community-dwelling subjects. A population-based study was conducted in 1992 and 1993, and subjects were re-investigated after 6 years. Plasma analytes were determined on both occasions. At follow-up, memory performance, using the Kendrick Object Learning Test, was investigated in 2,189 subjects (age, 65-67 years at baseline). Subjects with memory deficit (test score, < 25) had higher tHcy and lower folate at follow-up compared with those without memory deficit: 12.6 (95% confidence interval [CI], 12.1, 13.1) versus 11.5 (95% CI, 11.3, 11.6) micromol/L (p < 0.001) for tHcy, and 6.7 (95% CI, 6.2, 7.1) versus 7.6 (95% CI, 7.5, 7.8) nmol/L (p < 0.001) for folate. The risk of memory deficit increased according to quintiles of tHcy both at baseline and at follow-up. A decline in tHcy, or an increase in folate, over a 6-year period was associated with a higher memory test score; and vice versa. These findings indicate that increased plasma tHcy is an independent risk factor for memory deficit both cross-sectionally and prospectively, and that a "favorable" change in folate or tHcy concentrations over time is associated with better memory performance.