Developmental Pathways in EEG Activity in Infants at High Risk for Autism

Background: A current goal of research in autism is to examine the link between early markers and the later onset of the disorder.  Increasing attention is being paid to biomarkers of autism, including the brain’s electrical activity. Brain activity, as indexed by EEG, has been related to certain cognitive processes (Basar, 1999). Activity in the gamma band in particular is hypothesized to relate to weak coherence in perceptual processing, both of which are impaired in individuals with autism (Brock et al., 2002; Happe and Frith, 2006). This work has largely been done in older children and adults, but questions remain about the extent to which early differences in EEG activity, gamma as well as other frequency ranges, serve as early risk factors of autism and how well they predict later autism outcomes.

Objectives: This study represents a first step in examining the relationship between patterns of brain activity and behaviors associated with autism. We examine developmental trajectories in EEG activity in infants who are at high risk for developing autism and compare them to those of infants who are at low risk.

Methods: EEG is collected in the infants under an eyes open condition while a researcher blows bubbles to keep the infant still and quiet. We calculate power spectra for frequency bands that collectively span the 1-50 Hz range in infants at high risk for autism (by virtue of having at least one affected sibling) and compare them to infants at low risk (by virtue of having a typically developing sibling).  Infants are being studied at 6, 9, 12, and 18 months and this analysis is part of a larger, comprehensive infant-sibling project. Data are examined for patterns in developmental trajectories in various frequencies bands, each of which has a different functional relationship to cognitive processing.

Results: To date we have obtained usable data from 65 infants: 27 low risk and 38 high risk. Preliminary analysis indicates that high-risk infants exhibit lower power across all frequency bands at 6 months and that the subsequent development follows different paths. In particular, power in alpha activity at 18 months in high-risk infants is at the level of the 6 month alpha power in the low risk group suggesting that infants in the high-risk group may show a neurodevelopmental lag in alpha activity. For gamma activity, low risk controls show change over time in their power values while that of high-risk infants does not.

Conclusions: This analysis indicates that infants at high risk for autism demonstrate lower power in their EEG activity in each frequency range at six months of age. Additionally, the patterns of development in these neural measures show different patterns. Future studies will examine whether these developmental trajectories of power changes are associated with later autism behaviors.