OBJECTIVE: Moderately elevated levels of plasma total homocysteine are associated with an increased risk of developing Alzheimer's disease. We have tested whether baseline concentrations of homocysteine relate to the subsequent rate of cognitive decline in patients with established Alzheimer's disease (AD). METHODS: In 97 patients with AD, 73 pathologically-confirmed, we analysed the decline of global cognitive test scores (CAMCOG) over time from the first assessment for at least three 6-monthly visits up to a maximum of 9.5 years (in total 689 assessments). Non-linear mixed-effects statistical models were used. RESULTS: Baseline homocysteine levels showed a concentration-response relationship with the subsequent rate of decline in CAMCOG scores: the higher the homocysteine, the faster the decline. The relationship was significant in patients aged < 75 years who had not suffered a prior stroke. For example, in patients aged 65 years with a baseline homocysteine of 14 micromol/L, the decline from a CAMCOG score of 88 to a score of 44 occurred 19.2 (95% CI 6.8, 31.6) months earlier than in patients with a baseline homocysteine of 10 micromol/L. CONCLUSIONS: Raised homocysteine concentrations within the normal range among the elderly strongly relate to the rate of global cognitive decline in patients with Alzheimer disease. Plasma homocysteine can readily be lowered by B-vitamin treatment and trials should be carried out to see if such treatments can slow the rate of cognitive decline in relatively young patients with Alzheimer disease.
Int J Geriatr Psychiatry
82 - 90
Aged, Aged, 80 and over, Alzheimer Disease, Biomarkers, Cognition Disorders, Cohort Studies, Female, Homocysteine, Humans, Longitudinal Studies, Male, Neuropsychological Tests, Nonlinear Dynamics, Predictive Value of Tests