Homocysteine as a Biomarker for ADHD: A Systematic Review and Meta-Analysis with Protein-Protein Interaction Network Analysis

1.Attention Deficit Hyperactivity Disorder (ADHD) affects over 400 million individuals worldwide, yet reliable biomarkers for diagnosis remain elusive. Recent studies have suggested elevated homocysteine levels may serve as a potential biomarker, given its role in one-carbon metabolism and neurotransmitter synthesis. Objective is to conduct a comprehensive systematic review and meta-analysis examining homocysteine levels in ADHD patients compared to healthy controls, and to identify shared molecular pathways through protein-protein interaction network analysis. A systematic search was conducted across PubMed, Scopus, Web of Science, and Embase databases from inception to June 2025. Studies comparing plasma/serum homocysteine levels between ADHD patients and healthy controls were included. Meta-analysis was performed using random-effects models with standardized mean differences (SMD). Protein-protein interaction networks were constructed using STRING database for genes common to ADHD and hyperhomocysteinemia, with hub gene identification through CytoHubba analysis. Six studies comprising 796 ADHD patients and 488 controls from three countries were included. Meta-analysis revealed no statistically significant difference in homocysteine levels between groups (SMD = -0.38, 95% CI [-1.24, 0.48], p = 0.386), with substantial heterogeneity (I{superscript 2} = 85.4%). Results showed a biphasic distribution, with three studies demonstrating lower homocysteine in ADHD and two showing higher levels. Network analysis identified 485 common genes between ADHD and hyperhomocysteinemia, revealing 25 hub genes enriched in inflammatory pathways (TNF signaling), growth factor signaling (FGF family), and MAPK cascades. In conclusion homocysteine levels do not serve as a reliable standalone biomarker for ADHD diagnosis due to significant inter-study variability and population-specific factors. However, shared molecular networks suggest complex mechanistic relationships involving neuroinflammation, one-carbon metabolism, and neurotransmitter regulation. Future diagnostic approaches should consider multidimensional biomarker panels incorporating genetic, metabolic, and inflammatory markers rather than single metabolites.