Repetitive Behavior Symptoms Associated with Hippocampus and Amygdala Volumes in ASD

Background:  Although restricted and repetitive behaviors and interests (RRBI) are hallmark signs of autism spectrum disorders, the underlying neural mechanisms have received little attention in the literature. Recent behavioral research (e.g., Wigham et al., 2014; Rodgers et al., 2012) has suggested strong links between repetitive behaviors and symptoms of anxiety that are quite frequent in ASD samples. In typical populations, symptoms of anxiety are known to arise in part from atypical function of the medial temporal lobe (MTL) including amygdala and hippocampus. We therefore investigated whether MRI-obtained volumes of these structures may relate to anxiety and RRBI symptoms in ASD adults. 

Objectives: We hypothesized three-way associations between structural volumes of the hippocampus and amygdala with behavioral measures of RRBI and anxiety in our ASD sample more than in controls. 

Methods:   Young adults (ages 18-30) diagnosed with autism spectrum disorders (ASD, n =21) and age- and IQ-matched healthy college students (n = 20) completed MRI scans after completing a series of dimensional autism and emotion regulation surveys administered via internet-based software. Structural volumes of the hippocampus were obtained using hand-tracing by expert raters and additional ratings using established Freesurfer protocols. Amygdala volumes were obtained from Freesurfer. 

Results:  Hippocampus and amygdala volumes were not significantly correlated with either age or Full Scale IQ. In line with our previous work, the specific anxiety measure of intolerance of uncertainty was significantly correlated in the ASD with the Mannerisms subscale of the Social Responsiveness Scale-2^nd Edition (SRS-2); general trait measures of anxiety (the STAI-Trait and Beck Anxiety Inventory) were not. In the ASD group, both right and left hippocampus volumes (using both estimating methods) as well as amygdala volumes were significantly, positively correlated with the SRS-2 Mannerismsscale, but not with any anxiety measures. In contrast, the control group showed significant negative associations between trait anxiety and both left amygdala and left hippocampus. 

Conclusions:   The heterogeneity of autism spectrum disorders is challenging to match to brain-based measures, but narrower phenotypes may provide a useful platform for analysis. Here we show differential associations between brain volume of critical MTL structures with repetitive behaviors in ASD and trait anxiety in controls. Future research using both functional and other structural (e.g., DTI) methods may elucidate differences in these networks that aid understanding of etiology and targeted strategies for intervention.