Background: Autism Spectrum Disorder (ASD) is a neurodevelopment disorder characterized by socio-communicative impairments as one of the core symptoms. Autistic symptoms may be seen in the first year of life, they vary in severity from mild to severe, and in a few instances, they may improve over time, even without treatment. The neuropeptide N-acetyl-aspartyl-glutamate (NAAG) modulates glutamate release which has been proposed as a key mechanism underlying symptoms of ASD. NAAG provides one of the components of the proton magnetic resonance spectrum (1H-MRS) in humans. The signal of NAAG, however, largely overlaps with its precursor and degrading product N-acetyl aspartate (NAA) that by itself does not act in glutamatergic neurotransmission. Previously, we described the altered N-Acetyl-aspartyl-glutamate levels found in cingulated cortices by 1H-MRS in individuals with ASD that suggested neuronal damage. Taken together, the findings of this study support our hypothesis and a role for NAA-NAAG imbalance and impairments in the Social-communication skills in autism, which lead the next step in our investigation to correlate imbalances neurochemistry linked to cingulated cortices in social and communicational skills in the autism spectrum disorders.
Aim: To study the imbalance of NAA-NAAG metabolism role in the cingulated cortices correlated with the AQ domains social skills and communication associated with ASD severity using 1H-MRS.
Methods: We quantified NAAG and NAA separately from the 1H-MRS signal in 22 patients with ASD and 44 healthy comparison subjects, matched for age, gender on a 3.0 Tesla MR scanner. Autism quotients (AQ) scores were assessed. Statistic one-way ANOVA and Bonferroni correction was applied. Furthermore, the Pearson correlation hallmarks the goal.
Results: The results of the Pearson correlation were represented graphically, where it was observed that there is no correlation between the Socio-communicative skills and the NAA/NAAG ratio in the ACC (r = -0.43, P = .005) in the ASD group (Fig.1). However, when was stratified ASD plus TD groups as AQ1, AQ2, AQ3, and AQ4, there was within groups differences (AQ1, AQ2, AQ3, and AQ4) of NAA/NAAG ratio; was increased significantly (P = .05) in AQ3 (Fig. 2) and, decreased in AQ4. Comparably, there was no differences of (NAAG, NAA, or NAA/NAAG) concentrations in the PCC, but a positive linear correlation with communication (r = .55, P = .049) was observed. In addition, in both ACC and PCC, the AQ2, AQ3, and AQ4 groups maintain a different correlation pattern than the AQ1 group, both in social skills and communication showing the severity level change within AQ domains.
These results make us suggest the relation of the deficit socio-communicational with the enlarged relative grey matter volumes (rGMV) of the auditory network in ASD adults; in accordance with that described by (Watanabe & Rees, 2016); who demonstrated the relation of the deficits associated with the severity of autistic socio-communicational core symptom. Since NAA is considered a marker of neurons, these results provide stronger support for neuron loss in the posterior cingulated cortex than volume measurements by MRI alone.
Conclusion: We conclude that the concentrations of NAAG and NAA act differently in ASD. The opportunity to measure NAAG in subjects with ASD creates a new and promising approach for intensified research on the glutamatergic systems and on the effects of novel drug candidates.
Have you considered to seek out a Mahine Learning (ML) predictive model able to correlate NMR spectra and patients variables with the resulting Autism quotients (AQ) scores?
This could be a very interesting collaboration for our group or other group with interes in Biomolecular chem and biomedical engineering.
It could lead to the development of a computational user-friendly software useful in diagnosis, I guess, isn´t it?
Thank you very much for your valuable contribution to the article. Without a doubt, an appreciation very close to our objectives. Like you, we believe that collaboration with your group would respond to the prevailing need for an early clinical diagnosis for autism, from my point of view.
All kind regards,
Carmen Jimenez-Ezpinoza
Have been these results already published in a journal or only the early communication presented here?
Do you know if there previous Machine Learning studies on this area?
Thank you very much for your support
Please, feel to post your own questions to authors of other papers in this or other workshops to stimulate fruitful cross-over discussions.
Thank you very much for your interest.
These results only the early communication presented here.
I do not know the scope of Machine Learning in terms of Neurochemistry by 1H-MRS and autism, but I think that the previous Machine Learning studies have been developed with other tools but aren't with AQ-Test.
Thank you very much for your interest.
These results only the early communication presented here.
I do not know the scope of Machine Learning in terms of Neurochemistry by 1H-MRS and autism, but I think that the previous Machine Learning studies have been developed with other tools but aren't with AQ-Test.
Kind regards,
Carmen Jimenez Espinoza
