Loading...
Publication
Investigação de sinais eletrofisiológicos de processamento facial como biomarcadores na perturbação do espetro do autismo: Um estudo com arbaclofeno
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 4.81 MB | Adobe PDF |
Authors
Abstract(s)
A Perturbação do Espetro do Autismo (PEA) é uma condição do neurodesenvolvimento
complexa, caracterizada por défices na comunicação social e padrões de comportamento
restritos e repetitivos. Uma das hipóteses neurobiológicas centrais no autismo envolve
alterações do equilíbrio excitação/inibição (E/I) cerebral, no sentido de maior excitação. Este
estudo teve como objetivo avaliar o impacto do arbaclofeno, um agonista seletivo do recetor
inibitório GABA-B (ácido gama-aminobutírico) e modulador do equilíbrio E/I em adultos (idades
entre 19 e 52 anos) com e sem PEA durante uma tarefa de reconhecimento de faces. Para tal
foram utilizados os componentes N170 e P100 de eletroencefalografia (EEG) como
biomarcadores do processamento de informação facial.
Foram analisados dados de EEG de 70 participantes (29 com PEA e 41 controlos) num desenho
experimental duplamente cego, aleatorizado, cruzado e controlado por placebo. Cada
participante foi exposto a administração de três condições farmacológicas (placebo, 15 mg e 30
mg de arbaclofeno) durante uma tarefa de decisão percetual de deteção de faces. O préprocessamento
de sinais EEG foi automatizado, e os componentes P100 e N170 foram
identificados e quantificados. Os dados foram posteriormente submetidos a análises
estatísticas para investigar os efeitos do arbaclofeno nos marcadores eletrofisiológicos, e a
modelos de classificação baseados em algoritmos de machine learning para avaliar o potencial
discriminativo desses marcadores na diferenciação entre grupos diagnósticos e condições
experimentais.
As análises estatísticas revelaram que o arbaclofeno modulou a amplitude do P100 e N170 e a
latência do N170, com efeitos dose-dependentes e assimetrias hemisféricas. Notavelmente, o
fármaco prolongou a latência do N170 no hemisfério esquerdo exclusivamente no grupo de
controlo, sugerindo uma resposta diferenciada no grupo com PEA. Contudo, os algoritmos de
machine learning aplicados demonstraram capacidade discriminativa limitada para classificar
participantes por grupo diagnóstico, dose farmacológica ou resposta comportamental.
Os resultados indicam que, embora o arbaclofeno modifique as respostas eletrofisiológicas, a
elevada variabilidade interindividual e a sobreposição de dados limitam a utilização dos
componentes P100 e N170 como biomarcadores diagnósticos ou de resposta terapêutica
isolados na PEA. No entanto, os efeitos observados, nomeadamente as alterações dependentes
da dose e assimetrias hemisféricas, evidenciam a sensibilidade destes sinais eletrofisiológicos a
modulações farmacológicas e à sua potencial utilização como ferramenta para investigar os
mecanismos de ação de fármacos.
Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition characterised by deficits in social communication and restricted, repetitive patterns of behaviour. One of the central neurobiological hypotheses in autism involves alterations in the brain's excitation/inhibition (E/I) balance, in the sense of greater excitation. This study aimed to assess the impact of arbaclofen, a selective agonist of the inhibitory GABA-B (gamma-aminobutyric acid) receptor and modulator of the E/I balance in adults with and without ASD during a face recognition task. To this end, the N170 and P100 components of electroencephalography (EEG) were used as biomarkers of facial information processing. EEG data from 70 participants (29 with PEA and 41 controls) were analysed in a double-blind, randomised, crossover, placebo-controlled experimental design. Each participant was exposed to three pharmacological conditions (placebo, 15 mg and 30 mg of arbaclofen) during a perceptual decision task of face detection. EEG signal preprocessing was automated, and the P100 and N170 components were identified and quantified. The data were subsequently subjected to statistical analyses to investigate the effects of arbaclofen on electrophysiological markers, and to classification models based on machine learning algorithms to assess the discriminatory potential of these markers in differentiating between diagnostic groups and experimental conditions. Statistical analyses revealed that arbaclofen modulated the amplitude of P100 and N170 and the latency of N170, with dose-dependent effects and hemispheric asymmetries. Notably, the drug prolonged N170 latency in the left hemisphere exclusively in the control group, suggesting a differentiated response in the group with ASD. However, the applied machine learning algorithms demonstrated limited discriminatory ability to classify participants by diagnostic group, pharmacological dose, or behavioural response. The results indicate that, although arbaclofen modifies electrophysiological responses, high interindividual variability and data overlap limit the use of the P100 and N170 components as isolated diagnostic or therapeutic response biomarkers in ASDICA. However, the observed effects, namely dose-dependent changes and hemispheric asymmetries, demonstrate the sensitivity of these electrophysiological signals to pharmacological modulations and their potential use as a tool to investigate drug mechanisms of action.
Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition characterised by deficits in social communication and restricted, repetitive patterns of behaviour. One of the central neurobiological hypotheses in autism involves alterations in the brain's excitation/inhibition (E/I) balance, in the sense of greater excitation. This study aimed to assess the impact of arbaclofen, a selective agonist of the inhibitory GABA-B (gamma-aminobutyric acid) receptor and modulator of the E/I balance in adults with and without ASD during a face recognition task. To this end, the N170 and P100 components of electroencephalography (EEG) were used as biomarkers of facial information processing. EEG data from 70 participants (29 with PEA and 41 controls) were analysed in a double-blind, randomised, crossover, placebo-controlled experimental design. Each participant was exposed to three pharmacological conditions (placebo, 15 mg and 30 mg of arbaclofen) during a perceptual decision task of face detection. EEG signal preprocessing was automated, and the P100 and N170 components were identified and quantified. The data were subsequently subjected to statistical analyses to investigate the effects of arbaclofen on electrophysiological markers, and to classification models based on machine learning algorithms to assess the discriminatory potential of these markers in differentiating between diagnostic groups and experimental conditions. Statistical analyses revealed that arbaclofen modulated the amplitude of P100 and N170 and the latency of N170, with dose-dependent effects and hemispheric asymmetries. Notably, the drug prolonged N170 latency in the left hemisphere exclusively in the control group, suggesting a differentiated response in the group with ASD. However, the applied machine learning algorithms demonstrated limited discriminatory ability to classify participants by diagnostic group, pharmacological dose, or behavioural response. The results indicate that, although arbaclofen modifies electrophysiological responses, high interindividual variability and data overlap limit the use of the P100 and N170 components as isolated diagnostic or therapeutic response biomarkers in ASDICA. However, the observed effects, namely dose-dependent changes and hemispheric asymmetries, demonstrate the sensitivity of these electrophysiological signals to pharmacological modulations and their potential use as a tool to investigate drug mechanisms of action.
Description
Keywords
Perturbação do espetro do autismo (PEA) Eletroencefalografia (EEG) N170 Arbaclofeno Equilíbrio Excitação-Inibição Processamento facial Autism spectrum disorder (ASD) Electroencephalography (EEG) Arbaclofen Excitation-Inibition equlibrium Facial processing