Glutamate / Glutamine in the Basal Ganglia Is Associated with Executive Function and Communication Impairments in Autism: A [1H]MRS Study

Background: It has been suggested that dysfunction in glutamatergic neurotransmission may occur in Autism Spectrum Disorder (ASD). Glutamate is the brain’s primary excitatory neurotransmitter. However, few studies have measured brain glutamate in adults with ASD and related variation in glutamate to clinical phenotype. We previously [Lavender et al 2009 IMFAR poster, manuscript in preparation] reported that individuals with an ASD had reduced n-acetyl aspartate (NAA) and glutamate/glutamine (Glx) in the basal ganglia, as measured using proton magnetic resonance spectroscopy, [1H]MRS. However, it is not known whether these results are associated with particular symptoms of autism.

Objectives: We examined the correlation between basal ganglia glutamate/glutamine (Glx) concentration and clinical symptoms in terms of the three domains of the ADI-R diagnostic interview: Social impairment, Communication and language, and Repetitive interests and behaviours. We also measured executive function using the Zoo Task from the BADS neuropsychological battery since deficits in executive functioning are common in ASD(Hume et al. 2009) and interventions to improve executive functioning increase social independence (Hume et al. 2009).

Methods: We studied 31 adults with an ASD (4 female) diagnosed using ICD-10 criteria and the ADI-R interview. All participants had a normal range IQ > 70. The BADS Zoo Task, a measure of planning and problem-solving, involves designing a route through a fictional zoo, with the goal of visiting different exhibits, while adhering to a number of rules. [1H]MRS spectra were acquired on a 1.5T GE HDx MRI scanner (GE Medical Systems, Milwaukee, WI, USA). Single voxel spectra were acquired with a PRESS sequence with repetition time TR=3000ms, echo time TE=30 ms. Data were analyzed using LCModel software. MRS voxels were positioned in left medial parietal lobe (20x20x20 mm3), left dorsolateral prefrontal cortex (DLPFC) (16x24x20 mm3), and left basal ganglia (20x20x15 mm3) including parts of the caudate and putamen. T-tests showed no significant differences in the voxel ratios of grey matter to white matter between affected and control groups. As the results of the Zoo task were discrete (possible scores are 0,1,2,3, and 4), we used the nonparametric Spearman’s rank coefficient of correlation.

Results: Basal ganglia concentration of Glx concentration was significantly negatively correlated with scores on the ADI-R Communication scale. Lower i.e. more abnormal Glx was associated with greater communication impairment (r = -0.439, p = 0.020). On the Zoo task, there was likewise a negative correlation with basal ganglia Glx. The Spearman’s rank coefficient, rho=0.595 p=0.009 indicating a significant positive correlation between planning performance on the Zoo task, and basal ganglia combined glutamate/glutamine (“Glx”) levels. Lower i.e. more abnormal Glx was associated with worse executive function. By contrast, neither Social nor Repetitive symptom domains were correlated with basal ganglia Glx and no symptoms were correlated with Glx in dlPFC or parietal cortex (all p values >0.18). Zoo task performance was not correlated with Glx in the dlPFC (p=0.47) or parietal cortex (p=0.62).

Conclusions: Both executive functioning and communication and language impairment in ASD are associated with reduced basal ganglia glutamate/glutamine function.