Resting-State Neural Abnormalities in Autism Spectrum Disorders

Background: Previous research suggests neural abnormalities in individuals with Autism Spectrum Disorders (ASD), including aberrant activation in medial prefrontal cortex, posterior cingulate cortex, and the precuneus. These brain areas are implicated in social processing as well as in the “default mode network” that is active when participants are at rest. Although the default mode network is typically examined using fMRI, an investigation of resting-state neural oscillations may provide greater insight into the neural abnormalities in ASD. Whereas resting-state neural oscillations are aberrant in several neurological conditions and often associated with neurocognitive dysfunction, it is unclear whether similar associations manifest in ASD.

Objectives:  The present study probed neural oscillatory activity in nodes within the resting-state network in children with ASD. When differences between the ASD and control groups were observed, correlations between neural activity and clinical symptom severity were examined. It was hypothesized that children with ASD would show increased low-frequency activity, and that this increased activity would be associated with greater social and language impairments.

Methods: Participants were 18 children with ASD and 21 neurotypical control children who ranged in age from 6 to 14 years. All were medication-free. Whole-cortex MEG data were obtained while participants underwent a two-minute eyes-closed resting-state exam. Offline, a standard source model (15 regional sources) was used to transform each individual’s raw MEG surface activity into brain space. At each localized brain area, a Fast Fourier Transform (FFT) was applied to artifact-free two-second epochs of the continuous data. Individual spectra were averaged, and oscillatory activity was examined in 6 standard frequency bands: delta (0-4 Hz), theta (4-8 Hz), alpha (8-12 Hz), beta1 (12-20 Hz), beta2 (20-30 Hz), and gamma (30-50 Hz). Power in each frequency band (relative to total power) was compared between the ASD and control groups. When group differences were observed, correlations between relative power and scores on two clinical measures, the Social Responsiveness Scale (SRS) and the Clinical Evaluation of Language Fundamentals (CELF), were examined. 

Results: The ASD and control groups did not differ on total power (0-50 Hz). The ASD group, however, exhibited elevated relative frontal midline delta power (p = .02), as well as elevated relative alpha power at midline and right parietal sources (ps = .01 and .03, respectively). Frontal midline delta activity correlated with SRS (r = .32, p = .03) and CELF (r = -.31, p = .03) scores. Parietal alpha power correlated with SRS scores (rs = .36 and .31; ps = .01 and .03, for the midline and right hemisphere, respectively). 

Conclusions:  Results indicate abnormal resting-state brain activity in ASD. In particular, children with ASD exhibited greater frontal delta and parietal alpha power. Abnormalities in these regions appear to have clinical significance, as increased activity was associated with greater social and language impairments. These results provide converging evidence for atypical default mode network function in ASD, as the midline prefrontal and parietal regions investigated here correspond to nodes in the fMRI-identified default network.