Whole Brain Patterns of Cross-Frequency Coupling in ASD

Background: Phase amplitude coupling (PAC) has emerged as a fundamental resting-state mechanism by which low-frequency (4-12Hz) cortical oscillations modulate or gate higher frequency oscillations such as gamma-band activity (30-80Hz). PAC is thought to facilitate the input of long-range networks to local connections and subsequent processing. Abnormally high midline parietal-occipital resting-state alpha-gamma PAC has been previously observed in children with autism spectrum disorder (ASD). This study introduces a novel whole-brain mapping of PAC in order to elucidate the spatial pattern of resting-state cortical dysfunction in ASD.

Objectives: This study hypothesized that the PAC abnormalities in ASD are network-wide and pervasive.

Methods: Participants included 26 children with ASD (mean age = 9.5, SD = 2.2) and 22 typically developing (TD) children (mean age = 10.5, SD = 2.4). Five minutes of eyes-closed resting-state data were obtained using a 275-channel MEG system (VSM MedTech). Using BESA, a 15 regional source model assessed cortical activity across the entire brain at an approximately lobar level. At each cortical source, PAC was computed for each frequency pairing allowing whole-brain mapping of PAC (PAC-map) as a function of both space and frequency pairing. For this study, PAC between alpha (9 to 11 Hz) and gamma (30 to 40Hz) was examined.

Results: Across both ASD and controls, PAC decreased with age in midline frontal, left parietal and right parietal regions (p<0.05, each). In the ASD group, PAC was lower in midline frontal, left parietal, and right parietal regions (p<0.05 each) and elevated in the occipital-parietal midline region (p<0.05) compared to controls. In the midline frontal region, PAC decreased with age at a slower rate in ASD than controls (p<0.01).

Conclusions: PAC was visualized throughout the cerebral cortex to enable characterization of the development and heterogeneity of this vital neural network process. A primary finding was that PAC decreased in frontal and parietal regions as a function of age, with a slower frontal PAC decrease in ASD than controls. Whole brain PAC-map may form the foundation of future studies examining spatial patterns of cortical dysfunction in ASD.