Dysregulation of Temporal Dynamics of Neural Activity in Adolescents on Autism Spectrum

Background: Autism spectrum disorder is increasingly understood to be based on atypical signal transfer among multiple interconnected networks in the brain (Sun et al., 2012). Relative temporal patterns of neural activity have been shown to underlie both the altered neurophysiology and the altered behaviors in a variety of neurogenic disorders (Ahn et al., 2018). The present study investigated the temporal aspect of brain network stability and variability in resting state EEG of ASD adolescents and neurotypical controls.

Objectives: The purpose of the study is to assess brain network flexibility in ASD using measures of synchronization (phase-locking) strength (γ), and timing of synchronization and desynchronization of neural activity (desynchronization ratio, DR) in theta (4-7 Hz), alpha (8-12 Hz), beta (15-30 Hz), and low gamma (30-60 Hz) bands of resting state EEG.

Methods: Resting state EEG data was recorded from 14 individuals with the diagnosis of ASD (2 F, age M = 13.7, SD = 2.2, range 10–16) and 15 TD participants (5F, age M = 13.4 years, SD = 1.8, range 10–17). Pairwise γ and pairwise DR were computed between anterior (F3, F4) and parietal (P3, P4) electrode sites.

Results: EEG data of ASD participants manifested altered temporal of coordination in anterior and parietal brain regions in multiple frequency bands on very short temporal scales. Pairwise synchronization strength in fronto-parietal sites was, in general, lower in ASD (alpha P3-P4, p<.001; beta F3-P3, p<.001, F4-P4, p<.001; gamma F3-P3, p<.001, P3-P4, p<.012). Short desynchronizations were also more numerous in EEG data of ASD participants (alpha band: F3-P3 p<0.001, P3-P4 p<0.002; low gamma, F4-F4 p<.001, F4-P4 p<.046, P3-P4 p<.026). Pairwise analysis of right hemisphere electrodes (F4-P4) indicated an contrasting effect: synchronization was significantly lower in TD, as opposed to ASD participants, in the alpha band (alpha F4-P4, p<.005), while desynchronization ratio was higher in EEG data of TD participants over in the theta band (theta F4-P4, p<.04).

Conclusions: In ASD adolescents, stable cross-frequency networks during resting state have been previously shown to endure for a longer period of time, in comparison with the same in neurotypical peers (Malaia et al., 2016). The present analysis extends the findings to the temporal aspect of synchronization: we demonstrate that fronto-parietal synchronization is lowered in ASD, with more short periods of desynchronization. Mathematical modeling (Ahn & Rubchinsky, 2017) suggests that neural networks with high desynchronization ratio have increased sensitivity to inputs; this sensitivity may disrupt production of an adequate neural and behavioral responses to external stimuli. Cognitive processes dependent on integration of activity from multiple networks may be, as a result, particularly vulnerable to disruption. Contrasting findings of increased synchronization and lowered DR in the lower frequency bands over the right hemisphere in ASD participants require further exploration with regard to their relationship to linguistics (left-lateralized) vs. numeric (bilateral) manipulation skills. The study advances the understanding of the biological mechanisms underlying cortical computations in the complex ASD phenotype.