Barnes Samuel J K, Thomas Megan, McClintock Peter V E, Stefanovska Aneta
Department of Physics, Lancaster University, Lancaster LA1 4YB, UK.
Department of Paediatrics, Blackpool Teaching Hospitals NHS Foundation Trust, Blackpool FY3 8NR, UK.
Brain Commun. 2025 Feb 19;7(2):fcaf084. doi: 10.1093/braincomms/fcaf084. eCollection 2025.
Spontaneous electroencephalography (EEG) measurements have demonstrated putative variations in the neural connectivity of subjects with autism spectrum disorder, as compared to neurotypical individuals. However, the exact nature of these connectivity differences has remained unknown, a question that we now address. Resting-state, eyes-open EEG data were recorded over 20 min from a cohort of 13 males aged 3-5 years with autism spectrum disorder, and nine neurotypical individuals as a control group. We use time-localized, phase-based methods of data analysis, including wavelet phase coherence and dynamical Bayesian inference. Several 3 min signal segments were analysed to evaluate the reproducibility of the proposed measures. In the autism spectrum disorder cohort, we demonstrate a significant ( 0.05) reduction in functional connectivity strength across all frontal probe pairs. In addition, the percentage of time during which frontal regions were coupled was significantly reduced in the autism spectrum disorder group compared to the control group. These changes remained consistent across repeated measurements. To further validate the findings, an additional resting-state EEG dataset (eyes open and closed) from 67 individuals with autism spectrum disorder and 66 control group individuals (male, 5-15 years) was assessed. The functional connectivity results demonstrated a reduction in theta and alpha connectivity on a local, but not global, level. No association was found with age. The connectivity differences observed suggest the potential of theta and alpha connectivity as biomarkers for autism spectrum disorder. Additionally, the robustness to amplitude perturbations of the methods proposed here makes them particularly suitable for the clinical assessment of autism spectrum disorder and of the efficacy of therapeutic interventions.
与神经典型个体相比，自发脑电图（EEG）测量已证明自闭症谱系障碍患者的神经连接存在假定差异。然而，这些连接差异的确切性质仍然未知，我们现在要解决这个问题。对13名年龄在3至5岁的患有自闭症谱系障碍的男性和9名作为对照组的神经典型个体进行了20分钟的静息状态睁眼EEG数据记录。我们使用基于时间定位和相位的数据分析方法，包括小波相位相干和动态贝叶斯推理。分析了几个3分钟的信号段以评估所提出测量方法的可重复性。在自闭症谱系障碍队列中，我们证明所有额叶探测对之间的功能连接强度显著降低（P<0.05）。此外，与对照组相比，自闭症谱系障碍组额叶区域耦合的时间百分比显著降低。这些变化在重复测量中保持一致。为了进一步验证这些发现，对来自67名患有自闭症谱系障碍的个体和66名对照组个体（男性，5至15岁）的另外一组静息状态EEG数据集（睁眼和闭眼）进行了评估。功能连接结果表明，在局部而非全局水平上，θ波和α波连接减少。未发现与年龄有关联。观察到的连接差异表明，θ波和α波连接有潜力作为自闭症谱系障碍的生物标志物。此外，这里提出的方法对幅度扰动的稳健性使其特别适合用于自闭症谱系障碍的临床评估以及治疗干预效果的评估。
