Análise de Espectro ATR-FTIR de amostras de sangue e saliva de indivíduos com Transtorno do Espectro Autista

Abstract

Autism Spectrum Disorder (ASD) is characterized by a complex neurodevelopmental disorder, involving repetitive behaviors, speech impairment, restricted interests, and deficits in social skills to varying degrees. Its etiology is not fully elucidated, though genetic, environmental, and neurological factors are known to be implicated. ASD diagnosis is purely clinical and based on the criteria of the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-V), and the International Classification of Diseases (ICD-10), performed only by psychiatrists and neurologists. Early identification of ASD, therefore, favors a positive prognosis. Thus, this study aimed to develop a potential point-of-care method, based on ATR-FTIR spectroscopy, to aid in ASD diagnosis. Blood samples (plasma and serum) and saliva were collected from children with a clinical diagnosis of ASD (n=13), and a control group of neurotypical children (n=13) of both genders.Initially, 12 samples from each group underwent ATR-FTIR analysis to select the most appropriate clinical sample for the study. The data obtained were analyzed using ATR-FTIR, standard normal variate (SNV), principal component analysis (PCA), and leave-one-out cross-validation (LOOCV). Sample screening indicated no visible differences between plasma and serum across groups. Consequently, the study progressed with an increased sample size using saliva as the biological sample (n=13). Conversely, results demonstrated alterations in saliva composition among different study groups, correlating with spectrometric data. Machine learning algorithms provided an overall accuracy of 87.5% in classifying saliva data from children with ASD. Nonetheless, further studies with a larger sample size for external validation and different strategies for pre-sample preparation are necessary to improve the classification model.