Homocysteine and cognitive function in the elderly: the Rotterdam Scan Study.

BACKGROUND Elevated plasma total homocysteine (tHcy) concentrations are associated with AD and vascular dementia, but the relation with cognitive performance in nondemented elderly people is not known. OBJECTIVE To examine the association of tHcy and cognitive function in the elderly, and assess whether this may be mediated by structural brain changes on MRI. METHODS The Rotterdam Scan Study is a population-based study of 1,077 nondemented elderly. Cognitive performance was assessed, and compound scores were constructed for psychomotor speed, memory function, and global cognitive function. Cerebral infarcts, white matter lesions, and generalized brain atrophy were measured on MRI. The cross-sectional relationship between tHcy levels and neuropsychological test scores was assessed by multiple regression. RESULTS Mean tHcy level was 11.5 micro mol/L (SD 4.1). Increasing tHcy levels were associated with lower scores for psychomotor speed, memory function, and global cognitive function, and this was largely due to the association with tHcy levels in the upper quintile (>14 micro mol/L). Adjusted differences between test scores of participants in the upper quintile as compared with the lower four quintiles of tHcy were -0.26 (95% CI: -0.37; -0.14) for psychomotor speed, -0.13 (95% CI: -0.27; 0.01) for memory function, and -0.20 (95% CI: -0.30; -0.11) for global cognitive function. These associations were not mediated by structural brain changes on MRI. CONCLUSION Elevated tHcy levels are associated with decreased cognitive performance in nondemented elderly people, and the relation was most marked for psychomotor speed. This association was independent of structural brain changes on MRI.