White-matter lesions along the cholinergic tracts are related to cortical sources of EEG rhythms in amnesic mild cognitive impairment

Abstract

Does impairment of cholinergic systems represent an important factor in the development of amnesic mild cognitive impairment (aMCI), as a preclinical stage of Alzheimer's disease (AD)? Here we tested the hypothesis that electroencephalographic (EEG) rhythms, known to be modulated by the cholinergic system, may be particularly affected in aMCI patients with lesions along the cholinergic white-matter tracts. Eyes-closed resting EEG data were recorded in 28 healthy elderly (Nold) and 57 aMCI patients. Lesions along the cholinergic white-matter tracts were detected with fluid-attenuated inversion recovery sequences on magnetic resonance imaging. The estimation of the cholinergic lesion was performed with a validated semi-automatic algorithm pipeline after registration to a stereotactic template, image integration with stereotactic masks of the cholinergic tracts, and normalization to intracranial volume. The aMCI patients were divided into two groups of high (MCI Ch+; N = 29; MMSE = 26.2) and low cholinergic damage (MCI Ch-; N = 28; MMSE = 26.6). EEG rhythms of interest were delta (2-4 Hz), theta (4-8 Hz), alpha 1 (8-10.5 Hz), alpha 2 (10.5-13 Hz), beta 1 (13-20 Hz), and beta 2 (20-30 Hz). Cortical EEG generators were estimated by LORETA software. As main results, (i) power of occipital, parietal, temporal, and limbic alpha 1 sources was maximum in Nold, intermediate in MCI Ch-, and low in MCI Ch+ patients; (ii) the same trend was true in theta sources. These results are consistent with the hypothesis that damage to the cholinergic system is associated with alterations of EEG sources in aMCI subjects.

Figure 1

The trajectories of the cholinergic pathways superimposed onto the MRI scan of the template. (a) Midsagittal, (b) axial and (c) coronal views of the tracing of the pathways. The medial pathway, shown in green, originates from the nucleus basalis of Meynert and travels within the cingulum to supply the orbitofrontal, subcallosal, cingulate and retrosplenial cortices (a). The perisylvian lateral pathway courses within the claustrum to supply the opercular and insular cortices (b,c) and is represented in yellow. The capsular lateral pathway, shown in red, travels into the externale capsule and uncinate fasciculus (b,c) to supply the frontoparietal cortex, the middle and inferior temporal gyri, the inferotemporal cortex and the parahippocampal gyrus [Selden et al., 1998]. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Figure 2

Grand average of LORETA solutions (i.e. normalized relative current density at the cortical voxels) modeling the distributed EEG sources for delta, theta, alpha 1, alpha 2, beta 1, beta 2, and gamma bands in Nold, MCI C− (cholinergic lesion load <0.0065) and MCI Ch+ (cholinergic lesion load >0.0065). The left side of the maps (top view) corresponds to the left hemisphere. Legend: LORETA, low resolution brain electromagnetic tomography. Color scale: all power density estimates were scaled based on the averaged maximum value (i.e. alpha 1 power value of occipital region in Nold). The maximal value of power density is reported under each column. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Figure 3

Regional normalized LORETA solutions (mean across subjects) relative to a statistical ANOVA interaction (F(60,2460) = 5.44; MSe = 0.4; P < 0.00001) among the factors Group (Nold, MCI Ch−, MCI Ch+), Band (delta, theta, alpha 1, alpha 2, beta 1, beta 2, gamma), and ROI (central, frontal, parietal, occipital, temporal, limbic). This ANOVA design used the regional normalized LORETA solutions as a dependent variable. Subjects' age, education, gender and individual alpha frequency peak (IAF) were used as covariates. Regional normalized LORETA solutions modeled the EEG relative power spectra as revealed by a sort of “virtual” intracranial macro‐electrodes located on the macrocortical regions of interest. Legend: the rectangles indicate the cortical regions and frequency bands in which LORETA solutions presented statistically significant LORETA patterns Nold > MCI Ch− > MCI Ch+ (P < 0.05). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Figure 4

Regional normalized LORETA solutions (mean across subjects) relative to a statistical ANOVA interaction (F(30,1650) = 1.60; MSe = 0.2; P < 0.0216) among the factors Group (MCI Ch, MCI Ch+), Band (delta, theta, alpha 1, alpha 2, beta 1, beta 2, gamma), and ROI (central, frontal, parietal, occipital, temporal, limbic). This ANOVA design used the regional normalized LORETA solutions as a dependent variable. Subjects' age, education, gender and individual alpha frequency peak (IAF) were used as covariates. Regional normalized LORETA solutions modeled the EEG relative power spectra as revealed by a sort of “virtual” intracranial macro‐electrodes located on the macrocortical regions of interest. Legend: the rectangles indicate the cortical regions and frequency bands in which LORETA solutions presented statistically significant LORETA patterns MCI Ch− > MCI Ch+ (P < 0.05). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]