The profile of hippocampal metabolites differs between Alzheimer's disease and subcortical ischemic vascular dementia, as measured by proton magnetic resonance spectroscopy.

Alzheimer's disease (AD) and subcortical ischemic vascular dementia (SIVD) have overlapping pathologies and risk factors, but their underlying neurodegenerative mechanisms are basically different. We performed magnetic resonance spectroscopy (MRS) to study metabolite differences between the two diseases in vivo. The subjects were 31 patients with SIVD and 99 with AD. Additionally, 45 elderly subjects were recruited as controls. We measured N-acetylaspartate (NAA), glutamine and glutamate (Glx), and myoinositol (mIns) concentration quantitatively using a 1.5-T MR scanner. N-acetylaspartate and Glx concentrations decreased in the hippocampus and cingulate/precuneal cortices (PCC) in both AD and SIVD patients, and the NAA decrease in the hippocampus was more prominent in AD than in SIVD. Interestingly, the pattern of mIns concentration changes differed between the two disorders; mIns was increased in AD but not increased in SIVD. If one differentiates between AD and SIVD by the mIns concentration in the hippocampus, the area under the receiver operating characteristic curve was 0.95, suggesting a high potential for discrimination. Our results suggest that proton MRS can provide useful information to differentiate between AD and SIVD. The difference of mIns concentrations in the hippocampus and PCC seems to reflect the different neurodegenerative mechanisms of the two disorders.

Different atrophic patterns in early- and late-onset Alzheimer's disease and evaluation of clinical utility of a method of regional z-score analysis using voxel-based morphometry.

BACKGROUND/AIMS: We evaluated the differential patterns of brain atrophy in early- and late-onset Alzheimer's disease (AD) by measuring regional z-scores of voxel-based morphometry and assessed the availability of the method for clinical use. METHODS: The first 50 patients with probable AD were compared to 83 age-matched control subjects to identify the brain atrophy. Regions of interest were set in the areas showing z-scores >4. To find substantial differences in the atrophy pattern, principal component analysis was performed. The second group of 56 patients with memory complaints entered the study for evaluation of the clinical use of the model. RESULTS: The centers of the regions of interest were the amygdala, anterior hippocampi, posterior hippocampi, temporal cortices and subcallosal cortex, and left posterior cingulate cortex (PCC). Eigenvectors of the temporal cortices and left PCC showed counter-directions to those of patient age, suggesting that patients with younger onset age were preferentially associated with atrophy of those regions. Differential analyses of the second group showed high availability for the detection of abnormal brain atrophy in people with subjective memory complaints. CONCLUSION: AD with earlier onset is preferentially related to PCC and temporal lobe atrophy. Voxel-based morphometry can be statistically analyzed, and this method has the potential for bias-free assessment of brain atrophy.

Four subgroups of Alzheimer's disease based on patterns of atrophy using VBM and a unique pattern for early onset disease.

To clarify the involvement of the posterior cingulate cortex (PCC) in Alzheimer's disease (AD), we analyzed brain volume loss by voxel-based morphometry. Forty patients with non-familial AD and 20 patients with mild cognitive impairment (MCI) were recruited and compared to 88 elderly volunteers and 40 young volunteers. Local atrophy with aging was observed bilaterally in the perisylvian opercula, anterior cingulate cortex, caudate head, dorsomedial thalamus and parahippocampal cortex. In addition to those, atrophy in AD patients was observed in the amygdala, hippocampus, subcallosal region, posterior-associated cortices and PCC. We classified AD into four subgroups according to the atrophy pattern; atrophy in the amygdala/hippocampal formations (Hipp), in the Hipp and posterior cortices, in the Hipp and PCC and in the PCC and posterior cortices (PCC/-TOPa). As a result, the probability of PCC/-TOPa was 90% for ages <65 years, whereas that of the Hipp was 100% for disease duration >36 months. PCC atrophy was found in 16 of 40 AD patients and eight of 20 MCI patients. There seemed to be two subgroups with atrophy of the PCC, the one with disease progression and the other without. The latter had characteristic features of early onset and no significant atrophy in the amygdala/anterior hippocampus. There are at least four atrophy patterns that raise doubts about a single disease entity or progression in AD. This may reflect a different hierarchical pattern of progression in patients who have atrophy in PCC and posterior cortices when compared to the Braak staging scheme.