Fodor Zsuzsanna, Horváth András, Hidasi Zoltán, Gouw Alida A, Stam Cornelis J, Csukly Gábor
Department of Psychiatry and Psychotherapy, Semmelweis University, Budapest, Hungary.
Department of Neurology, National Institute of Clinical Neurosciences, Budapest, Hungary.
Front Aging Neurosci. 2021 Oct 7;13:680200. doi: 10.3389/fnagi.2021.680200. eCollection 2021.
While decreased alpha and beta-band functional connectivity (FC) and changes in network topology have been reported in Alzheimer's disease, it is not yet entirely known whether these differences can mark cognitive decline in the early stages of the disease. Our study aimed to analyze electroencephalography (EEG) FC and network differences in the alpha and beta frequency band during visuospatial memory maintenance between Mild Cognitive Impairment (MCI) patients and healthy elderly with subjective memory complaints. Functional connectivity and network structure of 17 MCI patients and 20 control participants were studied with 128-channel EEG during a visuospatial memory task with varying memory load. FC between EEG channels was measured by amplitude envelope correlation with leakage correction (AEC-c), while network analysis was performed by applying the Minimum Spanning Tree (MST) approach, which reconstructs the critical backbone of the original network. Memory load (increasing number of to-be-learned items) enhanced the mean AEC-c in the control group in both frequency bands. In contrast to that, after an initial increase, the MCI group showed significantly ( < 0.05) diminished FC in the alpha band in the highest memory load condition, while in the beta band this modulation was absent. Moreover, mean alpha and beta AEC-c correlated significantly with the size of medial temporal lobe structures in the entire sample. The network analysis revealed increased maximum degree, betweenness centrality, and degree divergence, and decreased diameter and eccentricity in the MCI group compared to the control group in both frequency bands independently of the memory load. This suggests a rerouted network in the MCI group with a more centralized topology and a more unequal traffic load distribution. Alpha- and beta-band FC measured by AEC-c correlates with cognitive load-related modulation, with subtle medial temporal lobe atrophy, and with the disruption of hippocampal fiber integrity in the earliest stages of cognitive decline. The more integrated network topology of the MCI group is in line with the "hub overload and failure" framework and might be part of a compensatory mechanism or a consequence of neural disinhibition.
虽然在阿尔茨海默病中已报道了α和β频段功能连接性（FC）降低以及网络拓扑结构的变化，但这些差异是否能标志该疾病早期的认知衰退尚不完全清楚。我们的研究旨在分析轻度认知障碍（MCI）患者和有主观记忆主诉的健康老年人在视觉空间记忆维持期间α和β频段的脑电图（EEG）FC及网络差异。在具有不同记忆负荷的视觉空间记忆任务期间，使用128通道脑电图对17名MCI患者和20名对照参与者的功能连接性和网络结构进行了研究。通过具有泄漏校正的幅度包络相关性（AEC-c）测量EEG通道之间的FC，而网络分析则通过应用最小生成树（MST）方法进行，该方法重建原始网络的关键主干。记忆负荷（待学习项目数量增加）在两个频段均增强了对照组的平均AEC-c。与此相反，在最初增加后，MCI组在最高记忆负荷条件下α频段的FC显著降低（<0.05），而在β频段则没有这种调节。此外，在整个样本中，平均α和β AEC-c与内侧颞叶结构的大小显著相关。网络分析显示，与对照组相比，MCI组在两个频段中，无论记忆负荷如何，最大度数、介数中心性和度数离散度均增加，直径和偏心率降低。这表明MCI组的网络重新布线，具有更集中的拓扑结构和更不平等的流量负荷分布。通过AEC-c测量的α和β频段FC与认知负荷相关的调节、内侧颞叶的细微萎缩以及认知衰退最早阶段海马纤维完整性的破坏相关。MCI组更整合的网络拓扑结构符合“枢纽过载和故障”框架，可能是补偿机制的一部分或神经去抑制的结果。
