Caudate Nucleus Volume And Restricted Repetitive Behaviors In2-to3-Year Old Children With Autism

[Objective]To evaluate caudate nucleus volume which is obtained from semi-automated segmentation and automated image analysis methods respectively in2-to3-yearold children with autism spectrum disorder (ASD) in comparison with children with developmental delay (DD), and to investigate the correlation between the abnormal change of caudate nucleus volume and severity of restricted, repetitive behavior (RRB). We also determine if similar results could be obtained from semi-automated and automated segmentation methods.[Method]Magnetic resonance imaging (MRI) scans using a three-dimensional SPGR imaging sequence were acquired in66ASD and36age, gender and development quotient (DQ) matched DD. Caudate nucleus structures were traced by semi-automated segmentation (ITK-SNAP) and automated image analysis (Freesurfer) respectively, and intracranial volume (ICV) were measured using Freesurfer. While intracranial volume (ICV) were measured using Freesurfer Software. The Caudate nucleus volume of ASD group and DD group were compared with or without adjustment for ICV. Besides,correlation between abnormal change of caudate nucleus volume and higher, lower order, total RSB score of ADI-R (Autism diagnostic interview-revised) and RBS-R (Repetitive behavior scale-revised) score in ASD group were examined individually. [Result]1) The comparisons of caudate volumes:Using semi-automated methods, the volumes of left (3504.15±508.68mm3vs3206.60±490.68mm3, P=0.005)、right (3683.31±483.59mm3vs3313.07±453.24mm3,P=0.001) and total (7187.46±959.00mm3vs6519.67±901.87mm3, P=0.001) caudate nucleus were significantly larger in the ASD group compared to the DD group. This difference in left (P<0.001)、right (P=0.001) and total (P<0.001) caudate nucleus volumes remained significant even after covarying for ICV.Using automated methods, the volumes of left(3423.95±411.67mm3vs3212.97±405.23mm3, P=0.015)、right(3543.98±396.30mm3vs3337.50±414.45mm3, P=0.015) and total (6967.939±795.16mm3vs6550.47±792.09mm3, P=0.013)caudate nucleus were significantly larger in the ASD group compared to the DD group. This difference in left (P<0.001)、right (P=0.001) and total (P<0.001) caudate nucleus volumes remained significant even after covarying for ICV.2) Correlation analysis:Pearson correlations were used to examine whether RRB measures were significantly associated with caudate nucleus volumes in the children with ASD. As shown in Table3, the volumes of left (r=0.359, P=0.004) and total0=0.317, P=0.013) caudate nucleus as measured by the semi-automated method were associated with ADI-C domain lower order (C3, C4) RRB scores. When intracranial volume (ICV) was entered as a covariate,left(r=0.417, P=0.001),Right (r=0.296, P=0.025) andtotal (r=0.372, P=0.004) caudate volumes as measured by semi-automated methods, left (r=0.384, P=0.003), right (r=0.363, P=0.006) and total (r=0.385, P=0.003) caudate volumes as measured by automated methods and were significantly correlated with lower order (C3, C4) RRB scores. These associations were not significant absent ICV adjustment..No significant relationships were found between ADI-C domain total RRB, ADI higher order RRB scores (C1, C2) or other subscales of the RBS-R and left, right or total caudate nucleus volume.In addition, left and total caudate nucleus volumes as measured by the automated method were significantly correlated with scores on the Restricted Interests subscale of the RBS-R.3) Internal consistency analysis of two methods:As shown in Fig.2, in these analyses, moderate linear relationships between these two methods existed for left, right and total caudate nucleus volumes. Additionally, intraclass correlations were calculated between two methods (left caudate=0.746, right caudate=0.706, total caudate=0.797), demonstrating a greater pattern of consistency.[Conclusion]Our findings revealed increased caudate nucleus volume in2-to3-year old children with ASD compared to DD controls, as well as an association between caudate volume and lower order RRB. This emphasizes the importance of the caudate nucleus in the pathophysiology of RRB in ASD. In addition, both semi-automated and automated methods of quantitative image analysis provided sensitive measures capable of detecting caudate nucleus volumes in2-to3-year old children with ASD and DD.