An Analysis of Resting EEG Data in Infants at High-Risk for Developing ASD

Thursday, May 12, 2016: 11:30 AM-1:30 PM
Hall A (Baltimore Convention Center)
T. Y. Kim, A. Naples, M. Rolison and J. McPartland, Child Study Center, Yale School of Medicine, New Haven, CT

Clinical diagnoses of autism spectrum disorder can be made as early as two years of age, but experimental evidence suggests that atypical patterns of attention and brain activity in infants who later develop autism may be present in the first months of life. Measurement of EEG in infancy provides a direct, non-invasive measure of cortical activity. Differences in EEG spectral power during rest have successfully discriminated children with ASD from controls and have been shown to correlate with clinical characteristics. Recent studies suggest that ASD can be characterized by atypical neural oscillations, such as decreased alpha power and increased theta power in adults relative to children. By measuring resting brain activity in infants at high (HR; older sibling with ASD) and normal risk (NR; no family history of ASD) for ASD, it may be possible to identify anomalous brain activity preceding emergence of behavioral symptoms.


The current study compared resting EEG spectral power in 6-24 month old HR and NR infants to assess (a) overall oscillatory power differences and (b) patterns of relative power within groups. We further assessed different patterns of age-related change in the EEG between groups.


In a longitudinal design, HR (n = 28, 46 sessions) and NR (n = 21, 31 sessions) infants were assessed at 6, 9, 12, and 24-months of age. EEG was recorded for 2 minutes with a 128-channel Hydrocel Geodesic Sensor net while infants observed bubbles blown by an experimenter. Data from electrodes over frontal scalp were processed with Netstation v4.5 through a first-order high-pass filter, a 100 Hz low-pass filter, and segmented into 120 1s epochs hand-edited for artifacts. The cleaned data was subjected to a Hann window and spectral power was estimated using a Fast Fourier Transform.


Preliminary analyses indicated that the ratio of alpha-to-theta (HR mean = -.4330, NR mean = -.3751, mean difference = .05787, p = .024) and the ratio of alpha-to-beta power (HR mean = .2671, NR mean = .3083, mean difference = .04121 in direction of NR, p = .017) differed between HR and NR infants. In NR infants, power in the beta frequency band was significantly correlated with age (r = .366, p = .043) but not in HR infants (r = .099, p = .513). Ongoing analyses explore patterns of coherence across scalp regions.


Results indicate different patterns of oscillatory brain activity with different maturational trajectories in HR versus NR infants. HR infants displayed decreased levels of alpha relative to both beta and theta activity. The lack of developmental change in beta activity in HR infants suggests atypical maturation of feedback connectivity in children at a higher risk of ASD. By identifying differences in neural oscillations in early life before behavioral symptoms emerge, differences in oscillatory power have the potential to serve as early biomarkers of ASD, as well as to provide insight into neural mechanisms of ASD that may inform novel forms of early intervention.