Front Psychiatry. 2026 Mar 30;17:1727227. doi: 10.3389/fpsyt.2026.1727227. eCollection 2026.
ABSTRACT
PURPOSE: To evaluate whether anti-glutamic acid decarboxylase (Anti-GAD) antibodies and selected biochemical parameters are associated with autism spectrum disorder (ASD), symptom severity, and autistic regression in a case-control design. We hypothesized that Anti-GAD titers would be higher in ASD and associated with greater symptom burden.
MATERIALS AND METHODS: Children aged 2-9 years with ASD diagnosed according to DSM-5 criteria and age- and sex-matched healthy controls were included. Laboratory analyses assessed Anti-GAD, ASO, ferritin, iron, vitamin D, vitamin B12, thyroid-stimulating hormone, and folate. Clinical evaluations included the Autism Behavior Checklist (ABC), Modified Checklist for Autism in Toddlers (M-CHAT), and Ankara Developmental Screening Inventory (AGTE). Autistic regression was defined as loss of previously acquired language and/or social communication skills lasting ≥3 months between 15 and 30 months of age.
RESULTS: Ninety children participated (45 with ASD and 45 controls). Anti-GAD antibody levels were higher in the ASD group than in controls (p = 0.003). Although statistically significant, the absolute difference between groups was modest. Among ASD cases, 62% met criteria for regression; this relatively high proportion may reflect the clinic-based sampling and the operational definition applied. Compared with those without regression, children with regression had higher Anti-GAD titers and ABC scores but lower iron and ferritin levels (all p < 0.05). Anti-GAD levels correlated positively with ABC total scores (r = 0.724, p = 0.01). These findings reflect statistical associations and do not imply causality.
CONCLUSION: Higher Anti-GAD antibody levels were statistically associated with ASD, autistic regression, and greater behavioral symptom severity in this sample. These results suggest possible immune-related contributions to ASD heterogeneity; however, larger longitudinal studies are needed before clinical or biomarker implications can be established.
PMID:41983249 | PMC:PMC13071506 | DOI:10.3389/fpsyt.2026.1727227
