Coupling Between Global and Regional Brain Structural Variation in Autism Is Modulated By Symptom Severity

Saturday, May 14, 2016: 11:30 AM-1:30 PM
Hall A (Baltimore Convention Center)
N. E. Foster1,2, M. Sharda1, K. A. R. Doyle-Thomas3, A. Tryfon1,2, E. Anagnostou3,4, A. C. Evans2, L. Zwaigenbaum5, J. D. Lewis2, J. P. Lerch6, K. L. Hyde1,2 and .. NeuroDevNet ASD Imaging Group7, (1)International Laboratory of Brain, Music and Sound Research, University of Montreal, Montreal, QC, Canada, (2)Montreal Neurological Institute, Montreal, QC, Canada, (3)Bloorview Research Institute, Holland Bloorview Kids Rehabilitation Hospital, Toronto, ON, Canada, (4)University of Toronto, Toronto, ON, Canada, (5)University of Alberta, Edmonton, AB, Canada, (6)Mouse Imaging Centre, Hospital for Sick Children, Toronto, ON, Canada, (7)NeuroDevNet, Vancouver, BC, Canada
Background: Brain imaging has revealed differences between individuals with Autism Spectrum Disorders (ASD) and Typical Development (TD), both in global (e.g., mean cortical thickness; CT) and regional structure. Global differences may reflect genetic or environmental influences acting across the brain in neurodevelopment, whereas local differences presumably reflect region-specific influences. However, global and regional factors in ASD have so far been studied separately, even though they act in concert to affect brain structure.

Objectives: 1) Examine how global CT drives regional CT differently in ASD, and 2) Test whether local uncoupling of CT in ASD is explained by symptom severity.

Methods: Participants were 48 ASD and 50 TD males from the NeuroDevNet project. Groups were age-matched (mean=12.8 years, SD=3.0) and had IQ>70. ASD participants were diagnosed using ADI-R and ADOS. CT was calculated from T1 MR images using CIVET software. Effects of age, IQ, site and brain volume were removed. Coupling between global mean CT and regional CT was assessed at two regional spatial levels: in 58 AAL regions (mean value), and by vertex (20mm smoothing). In both cases, the correlation of each region’s CT with global CT was calculated across subjects, separately by group. Group differences in correlation were assessed via Fisher transformation. For the AAL analysis, P values were corrected via false discovery rate. For vertex analysis, clusters were calculated using permutation testing to control familywise error (FWE) rate to P<0.05. Developmental effects were examined by splitting the sample at median age and repeating the analysis in both age cohorts. After identifying regions where global-regional correlation differed between groups, the contribution of each ASD subject to the correlation difference was quantified by factoring out the baseline TD relationship. The resulting residuals represented CT variation in ASD unaccounted for by the baseline global influence. Linear interaction analyses then tested modulation by symptom severity (SRS, ADI-R, ADOS) of the correlation between residual CT and global mean CT.

Results: A strong influence of global CT upon regional structure was evident in the global-regional CT correlations in both groups. Regional correlations were generally diminished in ASD vs TD. After FWE correction, a correlation decrease was found in right inferior frontal gyrus (IFG) in both AAL and vertex analyses. Younger and older participants showed similar effects. Interaction analyses showed that global-regional coupling in ASD was modulated by symptom severity (SRS) in the right IFG.

Conclusions: Variability in regional cortical structure is strongly driven by a global component in both TD and ASD, but there is a greater degree of variance unexplained by global CT in ASD. The more localized nature of structural variability in ASD is consistent with the idea that brain differences in ASD result from a complex interaction of genetic factors. These results are important from a functional perspective because of IFG’s role in social cognition and theory of mind, core features of ASD. This work offers a new approach to examine brain structural variability in special populations, and highlights the importance of accounting for global and local factors in brain structure.