Combined ROC curves for MRCC-I, MPGES, 8-isoprostane, and cPLA2 increase their discriminative value as biomarkers of autism spectrum disorders

Emerging evidence implicates mitochondrial dysfunction, oxidative stress, and neuroinflammation as key interrelated mechanisms underlying ASD pathophysiology. These molecular disturbances contribute to neuronal dysfunction through a self-reinforcing cycle, wherein mitochondrial impairment promotes reactive oxygen species generation, triggering lipid peroxidation and inflammatory signaling cascades in the brain. In the present study, the discriminative performance of mitochondrial respiratory chain complex I (MRCC-I), microsomal prostaglandin E synthase-1 (MPGES-1), 8-isoprostane, and cytosolic phospholipase A₂ (cPLA₂) as biomarkers for ASD was evaluated in 54 children with ASD and 37 age-matched healthy controls. Plasma levels of the selected biomarkers were quantified using enzyme-linked immunosorbent assays. Logistic regression was applied to assess the discriminative performance of combined biomarker panels through receiver operating characteristic (ROC) curve analysis. MRCC-I levels did not substantially change ( p = 0.881), whereas MPGES-1, 8-isoprostane, and cPLA₂ were significantly elevated ( p = 0.001) in ASD participants compared with controls. Individually, MPGES-1 (AUC = 0.870), 8-isoprostane (AUC = 0.815), and cPLA₂ (AUC = 0.889) demonstrated good discriminative performance, while MRCC-I alone showed limited utility (AUC = 0.510). Combined ROC analysis significantly improved discriminative accuracy, with multi-marker panels achieving an AUC of 0.976, sensitivity of 97.6%, and specificity of 100%. The combined use of MRCC-I, MPGES-1, 8-isoprostane, and cPLA₂ displays greater discriminative performance than individual markers, showing their potential as a multi-biomarker panel for the detection of ASD.