On the Spectrum

Chapter 5
Despite the presence of an extensive and expanding literature, the neurobiological etiology
of severe mental disorders such as autism spectrum disorder remains elusive. ASD has
traditionally been conceptualized categorically, but is increasingly recognized as the
severe end of a continuum of traits that extend into the general population (Constantino
and Todd 2003). Imaging studies using a phenotype of quantitative social impairment can
complement case-control studies to better understand the neurobiology of ASD. Irrespective
of classification approach, one of the prominent hypotheses on the origins of ASD is an
aberrant development of neuronal connections throughout the brain (i.e., ‘developmental
disconnection syndrome’, Geschwind and Levitt 2007).
Within this context, we utilized resting-state fMRI scans from a large, population-based cohort
study of children (Jaddoe, van Duijn, Franco, van der Heijden, van IIzendoorn et al. 2012,
White, El Marroun, Nijs, Schmidt, van der Lugt et al. 2013) to search for both underlying
maturational properties of dFNC that characterize age- and sex-specific connectivity
dynamics, and an underlying neurobiological substrate of autistic traits in the general
population. We hypothesized the presence of distinct dynamic connectivity states in children
that are similar to those already reported in adults, given many static networks are already
present at a young age (Gao, Gilmore et al. 2011). Further we expect age-related correlates of
dynamic connectivity to resemble adult-like patterns, where increasing age is associated with
states previously reported in adults. Lastly, as previous work has shown aberrant connectivity
dynamics in psychopathology, we hypothesize to see an association between aberrant
dynamic connectivity and features of autism. As traits of ASD have been shown to form a
continuum in the general population, we hypothesized that such connectivity features would
also be present along a continuum. Results showed that multiple brain domains (comprised
of sets of ICNs) that are widely recognized in studies of adults (e.g., sub-cortical, defaultmode
and sensorimotor) are also identified in this large group of young children. Results also
reveal that the dynamic properties of connectivity vary with both age and sex. Specifically,
we found increased inter-domain connectivity with age in the more mature, “adult-like”
dFNC states, in which older children also spent more time compared to the younger children.
Interestingly, children with autistic traits spent more time in a globally disconnected state,
which resembled the connectivity dynamics observed in younger children. These results show
a link between the typical and atypical development trajectories as captured by dynamic FNC,
where individuals with higher levels of autistic traits show a potentially delayed transition to
spending time in the globally modularized or more heavily connected states. Taken together,
the present study provides a conceptual framework to support further investigations of
typical and atypical brain development in the general population using novel neuroimaging
methodology and clinical insight.
100