Neurochemical and Neuroanatomical Changes in Children, Adolescents and Adults with Autism Spectrum Disorder

Background: Prefrontal cortex (PFC) is associated with executive functions, social cognition, and ability to adjust behavior. These characteristics are known to be impaired in patients with autism spectrum disorder (ASD). It has also been proposed that in ASD patients the PFC has local over-connectivity and long-range underconnectivity, which may be caused by increased overall excitation due to an imbalance in the excitation/inhibition ratio (E/I). These connectivity changes impair the integration of information flow from/to several cortical and subcortical brain regions compromising PFC functions, and brain systems to which it connects. Moreover, E/I imbalance may also play a role in abnormal neurodevelopment in ASD. 

Objectives: To investigate whether there is an imbalance in the E/I in the PFC of individuals with ASD, and whether cortical and subcortical volumetric differences are concomitantly present in ASD.  

Methods: ASD group:25 males; 15±5y.o (range:11-33); positive ADI-R and ADOS; FSIQ:93±14; typically developing (TD) group:22 males; 17±6y.o. (range:10-32); FSIQ:122±15; (mean±SD). All TD participants were screened with Social Communication Questionnaire and Social Responsiveness Scale excluding ASD symptomatology.Data were acquired in a 3T Siemens scanner. Two MPRAGE sequences were acquired for volumetric analysis. Freesurfer 5.0 was used to obtain cortical and subcortical volumes. A 3x3x3cm³ single-voxel located in the bilateral medial PFC was acquired using MEGA-PRESS for quantification of gamma-aminobutyric acid (GABA) (inhibition) and glutamate (excitation) ratios to N-acetylaspartate+N-Acetyl aspartylglutamic acid (NAA+NAAG). PRESS was used to obtain absolute brain metabolites concentrations: N-acetylaspartate (NAA), Choline (Cho), Inositol (Ins), and Creatine+phosphocreatine (tCr). Ratios and absolute concentrations were quantified with LCModel 6.3 1-D. Voxel tissue composition (gray matter, white matter, and cerebrospinal fluid) were measured using SPM and in-house developed software running in Matlab. Due to lack of quality 17 datasets were excluded from this particular analysis (9 ASD; 8 TD). Groups for neurochemical study were: ASD: 16 males, 15±3y.o (range:12-24; FSIQ:90±14; TD:14 males; 11-28 y.o., 18±6y.o (range:11-28); FSIQ:124±16 (mean±SD).  

Results: The Glu/NAA+NAAG was increased in the ASD group (p=0.01), whereas no changes were observed in GABA/NAA+NAAG, or other main brain metabolites. Volumetric measurements revealed decreased volumes in the left amygdala and right thalamus in the ASD group (p=0.009 and p=0.025, respectively); the left thalamus showed a tendency for reduced volume (p=0.06). Furthermore, in the ASD group, decreased volumes were detected in the left hemisphere: lateral orbitofrontal cortex (p=0.01), and pars orbitalis (p=0.01); as well as in the right insula (p=0.04). Increases were found in left paracentral gyrus (p=0.03), and right rostral anterior cingulate (p=0.01).  

Conclusions: Our results clearly point to an increase in E/I in ASD patients, however the relative nature of our measurements still leaves open the question on the source of such imbalance. We are acquiring unsuppressed water spectra for the absolute quantification of GABA and glutamate, which will enable us to clearly understand the direction of such imbalance.  Together with the neurochemical results, anatomical changes support thalamo-cortical deficits, as well as impaired systems involved in executive function and regulation of emotional processing and communication in ASD.