Hyperconnectivity in Dementia Is Early and Focal and Wanes with Progression.

We investigated in a longitudinal multicenter cohort study functional cortical connectivity changes along the course of frontotemporal dementia (FTD) and Alzheimer's disease (AD) from the prodromal stage of the diseases. Electroencephalography (EEG) was recorded in 18 FTD and 18 AD patients at the prodromal stage of dementia, at dementia onset, and 3 years after dementia onset. Twenty healthy controls (HC) underwent EEG recordings at the same time interval as the patients. Mutual information (MI) analysis measured the strength of functional network connectivity. FTD and AD patients showed greater MI at the prodromal stage of dementia (FTD vs. HC P = 2 × 10-8; AD vs. HC P = 4 × 10-3). Local connectivity was higher in left and right frontal areas of FTD (P = 7 × 10-5 and 0.03) and in left and right posterior areas in AD (P = 3 × 10-5 and 5 × 10-5) versus HC. We showed cortical hyperconnectivity at the prodromal stage of dementia in areas involved in the specific pathological process of FTD (frontal regions) and AD (posterior regions). Hyperconnectivity disappeared during follow-up, thus suggesting that it is an early electrophysiological feature of dementia, potentially useful to identify prodromal FTD and AD.

Anterior EEG slowing in dementia with Lewy bodies: a multicenter European cohort study.

Electroencephalography (EEG) slowing with prealpha dominant frequency (DF) in posterior derivations is a biomarker for dementia with Lewy bodies (DLB) diagnosis, in contrast with Alzheimer's disease (AD). However, an intrasubject re-evaluation of the original data, which contributed to the identification of EEG DLB biomarker, showed that DF was slower in anterior than posterior derivations. We suppose this anterior-posterior gradient of DF slowing could arise in DLB from a thalamocortical dysrhythmia, differently involving the anterior and posterior cortical areas, and correlating with cognitive impairment (Mini-Mental State Examination). EEG was recorded in 144 DLB, 116 AD, and 65 controls from 7 Centers of the European DLB Consortium. Spectra were divided into delta, theta, prealpha, alpha frequency bands. In DLB, mean DF was prealpha both anteriorly and posteriorly, but lower anteriorly (p < 0.001). In 14% of DLB, DF was prealpha anteriorly, whereas alpha posteriorly. In AD and controls, DF was constantly alpha. EEG slowing in DLB correlated with cognitive impairment. Thalamocortical dysrhythmia gives rise to prealpha rhythm with an anterior-posterior gradient and correlates with impaired cognition.

Cortical Network Topology in Prodromal and Mild Dementia Due to Alzheimer's Disease: Graph Theory Applied to Resting State EEG.

Graph theory analysis on resting state electroencephalographic rhythms disclosed topological properties of cerebral network. In Alzheimer's disease (AD) patients, this approach showed mixed results. Granger causality matrices were used as input to the graph theory allowing to estimate the strength and the direction of information transfer between electrode pairs. The number of edges (degree), the number of inward edges (in-degree), of outgoing edges (out-degree) were statistically compared among healthy controls, patients with mild cognitive impairment due to AD (AD-MCI) and AD patients with mild dementia (ADD) to evaluate if degree abnormality could involve low and/or high degree vertices, the so called hubs, in both prodromal and over dementia stage. Clustering coefficient and local efficiency were evaluated as measures of network segregation, path length and global efficiency as measures of integration, the assortativity coefficient as a measure of resilience. Degree, in-degree and out-degree values were lower in AD-MCI and ADD than the control group for non-hubs and hubs vertices. The number of edges was preserved for frontal electrodes, where patients' groups showed an additional hub in F3. Clustering coefficient was lower in ADD compared with AD-MCI in the right occipital electrode, and it was positively correlated with mini mental state examination. Local and global efficiency values were lower in patients' than control groups. Our results show that the topology of the network is altered in AD patients also in its prodromal stage, begins with the reduction of the number of edges and the loss of the local and global efficiency.

Default network is not hypoactive in dementia with fluctuating cognition: an Alzheimer disease/dementia with Lewy bodies comparison.

Default mode network resting state activity in posterior cingulate cortex is abnormally reduced in Alzheimer disease (AD) patients. Fluctuating cognition and electroencephalogram abnormalities are established core and supportive elements respectively for the diagnosis of dementia with Lewy bodies (DLB). Our aim was to assess whether patients with DLB with both of these features have different default mode network patterns during resting state functional magnetic resonance imaging compared with AD. Eighteen patients with DLB, 18 AD patients without fluctuating cognition, and 15 control subjects were selected after appropriate matching and followed for 2-5 years to confirm diagnosis. Independent component analysis with functional connectivity (FC) and Granger causality approaches were applied to isolate and characterize resting state networks. FC was reduced in AD and DLB patients compared with control subjects. Posterior cingulate cortex activity was lower in AD than in control subjects and DLB patients (p < 0.05). Right hemisphere FC was reduced in DLB patients in comparison with control subjects but not in patients with AD, and was correlated with severity of fluctuations (ρ = -0.69; p < 0.01). Causal flow analysis showed differences between patients with DLB and AD and control subjects.