Skip to content
Review Open access

Biomarkers for the diagnosis of autism spectrum disorder: a systematic review of diagnostic accuracy and clinical validity

Oct 2026 · ANPAbstract_CBN26 · 0 citations

Abstract

Background: Autism Spectrum Disorder (ASD) presents a high global prevalence, and its diagnosis remains strictly clinical, based on behavioral scales. Given the etiological heterogeneity and the absence of consolidated biological tests, this systematic review analyzes the diagnostic accuracy and clinical validity of potential biological biomarkers to optimize early identification. Objective: To evaluate the diagnostic accuracy and clinical validity of biological biomarkers for the early identification of Autism Spectrum Disorder (ASD). Methods: A systematic review was conducted according to PRISMA 2020 guidelines, with the protocol submitted to PROSPERO (under review), structured in two stages. The search in PubMed, Embase, Scopus, and SciELO identified 432 records; after removing 96 duplicates, 336 were screened by title and abstract. Twenty-three studies progressed to full-text reading in Rayyan, and 8 were included in the final qualitative synthesis. Methodological quality was assessed using the QUADAS-2 tool, evidencing a predominantly moderate risk of bias, primarily related to the case-control design. Results: Analysis of the 8 selected studies (2024–2026), involving 59 to 313 preschool children, revealed biomarkers across four main domains, all validated against gold standards (DSM-5, ADOS-2, or ADI-R). Structural and functional neuroimaging showed the highest accuracy (AUC up to 0.97), with connectivity alterations in frontal regions and basal ganglia correlating directly with clinical severity. In the molecular field, proteomic panels and the CTRP1 protein indicated an early systemic inflammatory signature (sensitivity ~76%), suggesting extra-cerebral pathophysiological pathways. The metabolic convergence observed in the gut-brain axis (GABA/glutamate ratio and BCAAs, AUC 0.756) reinforces the multidimensional nature of ASD, while the integration of multimodal neurophysiological markers (EEG and eye-tracking) demonstrated the highest sensitivity (>90%) for identifying core symptoms. These findings indicate that biomarkers can provide the objectivity needed to mitigate uncertainties in strictly behavioral diagnoses. However, clinical validity for large-scale implementation remains limited by methodological heterogeneity, moderate sample sizes, and potential publication bias. The dependence on clinical instruments as the gold standard and the need for operational cost reduction remain critical barriers to a definitive transition toward precision medicine in child neurology. Conclusion: These findings suggest that ASD diagnosis is evolving from a purely behavioral framework toward precision medicine. This transition is crucial for facilitating intervention within the window of peak neuroplasticity, although clinical implementation still requires cost simplification and large-scale validation. Furthermore, clinical evaluation based on traditional diagnostic methods remains essential to the current diagnostic process.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.