Domain-Specific Effects of GABA-Modulating Pharmacotherapies in Autism Spectrum Disorder: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Background: Altered gamma-aminobutyric acid (GABA) neurotransmission is a proposed mechanism underlying autism spectrum disorder (ASD), prompting evaluation of several GABA-modulating pharmacotherapies. However, it remains unclear whether these interventions improve ASD broadly or preferentially affect specific symptom domains. Methods: We conducted a systematic review and random-effects meta-analysis of randomized controlled trials evaluating GABA-modulating pharmacotherapies in individuals with ASD. PubMed/MEDLINE, Embase, Scopus, and CENTRAL were searched from inception to April 1, 2026. Outcomes were prespecified as global autism severity, social communication, functional communication, restricted and repetitive behaviours (RRBs), adaptive behaviour, and irritability. Risk of bias was assessed using the Cochrane Risk of Bias 2 tool, and certainty of evidence was evaluated using GRADE. Results: Thirteen randomized controlled trials evaluating three GABA-modulating interventions (bumetanide, arbaclofen, and valproate) were included. GABA-modulating therapies were associated with statistically significant improvements in global autism severity (Hedges' g = -0.25, 95% CI -0.46 to -0.03; p = 0.028) and adaptive behaviour (Hedges' g = -0.09, 95% CI -0.15 to -0.02; p = 0.023). No significant pooled effects were observed for social communication (Hedges' g = -0.26, p = 0.077), functional communication (Hedges' g = -0.01, p = 0.869), RRBs (Hedges' g = -0.21, p = 0.126), or irritability (Hedges' g = -0.07, p = 0.543). After Holm-Bonferroni step down procedure, neither global autism severity nor adaptive behaviour remained statistically significant (Holm-adjusted p = .140 and .138, respectively). Adverse events were predominantly gastrointestinal, neurological, metabolic, and appetite-related. Overall risk of bias was variable, and the certainty of evidence ranged from very low to moderate. Conclusions: GABA-modulating pharmacotherapies did not demonstrate a robust, multiplicity-corrected benefit in any of the six prespecified ASD symptom domains. Nominal, unadjusted improvements in global autism severity and adaptive behaviour did not withstand correction for multiple comparisons and should be regarded as hypothesis-generating rather than confirmatory. Larger, adequately powered randomized trials using standardized domain-specific outcome measures are needed to determine whether individual GABA-modulating agents provide clinically meaningful benefit.