Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder. Despite its biological underpinnings, ASD is yet diagnosed based on behavioral assessments. The importance of early ASD detection is supported by a large body of evidence. Early diagnosis can facilitate more effective interventions, reducing the socioeconomic costs and improving the quality of life of both children with ASD and their families. This study aims to investigate the expression patterns of ASD-related microRNAs (miRNAs) in Iranian children with ASD compared with matched neurotypical controls, to identify candidate miRNA biomarkers.

Validation of ASD-related miRNAs was performed using the qRT-PCR method in plasma samples from 20 ASD and 20 unaffected individuals. Differentially expressed miRNAs were identified based on relative quantification analysis. Then, the diagnostic potential of each candidate miRNA was assessed using receiver operating characteristics (ROC) curve analysis.

Based on statistical and bioinformatics analyses, eight miRNAs were recommended as candidate biomarkers. Furthermore, differential expression of five of these miRNAs—miR-146a-5p, miR-338-3p, miR-181b-5p, let-7a-5p, and miR-140-3p—was confirmed experimentally by qRT-PCR (p-value < 0.05) based on a t-test. miR-140-3p was not further detected as significant by the Mann–Whitney test. The sensitivity, specificity, and area under the ROC curve were calculated for each validated miRNA and fell in the ranges of 75%–91%, 66.67%–83.33%, and 74%–90%, respectively.

This study revealed a panel of high-confidence differentially expressed miRNAs, which requires further investigation in larger sample sizes and by other validation tests. These findings could be a step forward in the field of ASD biomarker discovery.