Dynamic modulation of frontal theta at 10 months predicts cognitive development at age 3: effects in infants with and without a family history of autism

• In Event: The theta rhythm as a neural signature of early learning processes

Abstract

Background An increase in oscillatory power in the theta band on presentation of a novel stimulus may reflect learning in infancy (Begus et al., 2020). indeed, individual differences in dynamic theta responses to novel stimuli at 12-14 months predict cognitive development in later childhood (Jones et al., 2020). Understanding the robustness of these associations could illuminate the mechanisms through which early individual differences in infant learning relate to longer term cognitive outcomes. Here, we test whether the predictive validity of theta modulations extends to earlier in infancy (within the first year of life); whether long-term associations with later cognitive development are mediated through earlier developmental skills; and how these processes differ in infants with a family history of neurodevelopmental disorders.
Methods Participants were a cohort of n=136 infants with and without older siblings with a clinical diagnosis of ASD and/or ADHD who provided sufficient artifact-free EEG data at 10 months. Cognitive data was collected using the Mullen Scales of Early Learning (verbal and nonverbal domains) at 10, 14, 24 and 36 months. EEG was collected using a 128-channel Geodesic sensor net while the infant viewed novel social dynamic videos (duration = 1 minute). Continuous EEG was filtered (.1-100Hz), divided into one-second segments and artifacts were removed. Artifact-free segments were fourier transformed and mean power in the theta range (3-6Hz) was extracted for the first and second halves of the first video presentation (following Jones et al., 2020).
Results: Replicating previous work, a mixed model indicated a significant relation between frontal theta change at 10 months and cognitive scores at 36 months (t(98)=2.40, p = 0.018; Figure 1); this was significantly stronger in infants without a family history of ASD (t(98)=-2.21, p = 0.029). There were no significant associations with cognitive development at 10 (t(126)=0.01, p = 0.9), 14 (t(126)=-1.04, p = 0.3) or 24 months (t(106)=1.16, p = 0.25). Longitudinal data was then entered into a structural equation model using maximum likelihood estimation (lavaan). Model estimates indicated increasing strength of association from theta change to cognitive development over time (Figure 2), with a significant link at 36 months for the infants without a family history of ASD, (est=0.075, z=2.22, p = 0.027) but not those who did have a family history of ASD (est=-0.002, z=-0.059, p = 0.95).
Discussion These results are consistent with previous work in showing a relation between dynamic modulation of frontal theta in response to novel experiences in infancy and later cognitive development (Jones et al., 2020, Braithwaite et al., 2020). We show that these associations are present from earlier in infancy than previously demonstrated, but not in those with elevated risk of developing ASD. Finally, the association between early theta rhythms and cognition becomes stronger when cognition is measured later in development.
Further work should examine whether this reflects the increasing accuracy and heritability of behavioural measures of cognitive development over this time period. Overall, individual differences in theta modulation in response to new information may be sensitive to variation in early learning capacity.

Group Authors

STAARS team

Authors

• Emily J.H. Jones, Birkbeck, University of London
  Presenting Author

• Jannath Begum Ali, Centre for Brain and Cognitive Development
  Non-Presenting Author

• Amy Goodwin, King’s College London
  Non-Presenting Author

• Tony Charman, King's College London
  Non-Presenting Author

• Mark H. Johnson, University of Cambridge
  Non-Presenting Author