Objectives: Using whole brain gray-white brain segmentation, we sought to identify total and regional brain volume and asymmetry differences between ASD and typically developing (TD) subjects.
Methods: MRI scans were acquired on a 3T Siemens, and segmentation data was analyzed for volume differences between groups, as well as for regional asymmetries and asymmetry differences between groups. Participants included 10 ASD (8.5 +/-1.27y) and 11 TD (8.55 +/-1.97y). The analyses were performed on both raw data (non-normalized) as well as after normalizing for cerebral volume. Effect sizes were calculated [(Mean volume for autistics-mean volume for typical)/Pooled Standard deviation] on right side, left side and bilaterally (right+left) for all segmentation volumes. Then, mean ASD and TD segmentation volumes were compared between groups using the Multivariate ANOVA for correlated data (Proc Mixed in SAS) to obtain significance (p) values for each of the segmentation volume. Finally, Symmetry Indices (SI) [2*((Volume Left-Volume Right)/(Volume Left+Volume Right))*100] were calculated and compared between groups. Also, a one-sample t-test was performed to test whether the SIs in each group were significantly different from 0. All between-group analyses included adjustments for age, handedness and non-verbal IQ.
Results: Total cerebral volume asymmetry was significant for autistics (t -2.39; p 0.0406) but not for TD (t -0.98; p 0.3496). However there were no significant between group differences in either total cerebral volume (F 0.04; p 0.85) or total cerebral volume asymmetry (F 3.00; p 0.1090). Analysis of the data normalized to cerebral volume showed that Caudate was significantly smaller in ASD than TD on right (ES -0.68; p 0.0316), left (ES -0.787; p 0.047) and bilaterally (ES -0.77; p 0.03), although there were no differences in raw caudate volume between ASD and TD groups. Total Amygdala volume adjusted for cerebral volume was smaller in ASD than TD (p=.0375, ES = -.43) but neither right nor left nor any raw volumes of amygdala were different between groups. Significant asymmetries were found in both ASD and TD in putamen (ASD: p=.0015; TD:p=.0002) and pallidus (ASD: p=.0011,TD: p=.0008) and in TD only for cerebellar white matter (p=.0495) and amygdala (p=.0371).
Conclusions: Our preliminary findings are consistent with other studies showing differences in caudate and amygdala between ASD and TD subjects. We also note a modest loss of regional asymmetry in ASD. Given the lack of difference in total brain volume between groups, the source of differences in significance of amygdala and caudate measures between raw and adjusted volumes remains unclear.