Plasma EAAT2 and GABA as candidate biomarkers in males with autism spectrum disorder: an exploratory case–control study with ROC analysis

Autism spectrum disorder (ASD) lacks validated laboratory markers for diagnosis or risk stratification. Astrocyte-linked glutamatergic regulation and GABAergic signaling may be altered in ASD. This exploratory study quantified plasma excitatory amino acid transporter 2 (EAAT2), γ-aminobutyric acid (GABA), and the GABA_A receptor α5 subunit (GABRA5) and described their apparent discriminative performance. In a case-control sample of 46 male participants with ASD and 26 age-matched typically developing male controls, plasma biomarkers were measured by ELISA. Group differences were assessed with nonparametric tests. Receiver operating characteristic (ROC) analyses summarized discrimination. Logistic regression was used solely to generate combined-marker ROC curves; coefficients were not interpreted. EAAT2 and GABA concentrations were lower in ASD than controls (both P < 0.001). GABRA5 was directionally lower but not statistically significant (P = 0.149). Individual AUCs were 0.952 for EAAT2, 0.827 for GABA, and 0.646 for GABRA5. Combined-marker models yielded high apparent AUCs (EAAT2 + GABA+GABRA5: 0.993) within this dataset. Lower plasma EAAT2 and GABA are associated with ASD status in this male sample. Combined markers show high apparent discrimination in a case-control design, but estimates are optimism-biased and not diagnostic. Replication with comprehensive assay validation, calibration, and clinically representative cohorts (including females and relevant differentials) is required.