Dynamic Whole-Brain Functional Connectivity and Connectopathy in Autism: A Population-Based Neuroimaging Study of Brain Development

Thursday, May 12, 2016: 11:30 AM-1:30 PM
Hall A (Baltimore Convention Center)
R. L. Muetzel1, L. M. Blanken1, B. Rashid2, R. Miller3, F. Verhulst4, H. Tiemeier4, V. Calhoun3 and T. J. White5, (1)Child and Adolescent Psychiatry, Erasmus Medical Center - Sophia Children's Hospital, Rotterdam, Netherlands, (2)Electrical and Computer Engineering, University of New Mexico, Albuquerque, NM, (3)University of New Mexico, Albuquerque, NM, (4)Department of Child and Adolescent Psychiatry/Psychology, Erasmus MC-Sophia, Rotterdam, Netherlands, (5)Erasmus University Medical Centre, Rotterdam, Netherlands
Background:  Deficits in brain connectivity have been proposed to underlie symptomatology in autism spectrum disorders (ASD). Functional brain connectivity has been extensively studied in children and adults with ASD since the introduction of resting-state functional MRI (RS-FMRI). Despite the presence of an extensive and expanding literature, both in terms of brain function and structure, the neurobiological etiology of ASD remains elusive. Traditional connectivity studies operate under the assumption that the characteristics of connectivity remain stationary throughout the measurement period. Dynamic functional connectivity (DFC), a new technique that relaxes this stationarity assumption and allows states to change over time, may show promise in the study of connectopathy in children with ASD.

Objectives:  ASD has traditionally been conceptualized categorically, but is increasingly recognized as the end of a continuum of traits that extends into the general population(1). While these two constructs of the disorder clearly compliment one another in the search for the underlying neurobiology, there is an overrepresentation of case-control studies in the literature. The current study aims to expand the current functional connectivity literature in ASD in two ways: First by using a dimensional measure of autistic traits and second by using a novel metric of functional connectivity, namely DFC, in both globally ‘disconnected’ and default-mode states.

Methods:  The present work utilizes a large, population-based cohort of children ages 6-to-10 years (mean age = 7.9 years). 774 children (52% boys / 48% girls) participated in the study. Mean non-verbal IQ was 102±14. Autistic traits were measured using the short form (18 item) Social Responsiveness Scale. Children underwent a 5½-minute RS-FMRI scan with a 3 Tesla GE MRI system. MRI data were pre-processed using the Statistical and Parametric Mapping Software, and group independent component analysis was performed using the GIFT-toolbox. Special care was taken to account for movement artifact(2). Traditional static connectivity was first estimated, followed by DFC using a sliding-window approach and k-means clustering(3). Summary metrics, including mean dwell time were computed for each participant and dynamic state. Multiple linear regression analyses were conducted to assess the association between DFC and autistic traits.

Results:  Figure 1A illustrates the average, static connectivity state, and the 4 dynamic connectivity states observed in the current sample. Linear regression analyses, accounting for age, sex, and non-verbal IQ, show that autistic traits are positively associated with average time spent in a globally disconnected state (state 2, Figure 1B). Conversely, low levels of autistic traits were associated with a greater average time spent in a default-mode state.

Conclusions:   For the first time, we demonstrate associations between autistic traits on a continuum and aspects of DFC in a large, population-based study of children. Specifically, we show evidence of children with high levels of autistic traits spending more time in a globally ‘disconnected’ functional state, relative to children with lower levels of autistic traits.


1. Constantino, J.N., Todd, R.D., 2003. Arch Gen Psychiatry 60: 524-30.

2. Power, J.D., et al. 2012. Neuroimage 59: 2142-54.

3. Allen, E.A., et al. 2014. Cereb Cortex 24: 663-76