Can bumetanide be a miraculous medicine for autism spectrum disorder: Meta-analysis evidence from randomized controlled trials

The efficacy of bumetanide on autism spectrum disorders (ASD) has been extensively studied, but the results remain inconclusive, necessitating a critical examination to determine its therapeutic role. Therefore, this study conducted a meta-analysis of bumetanide treatments for ASD to explore its efficacy and identify subpopulation that responded positively. A systematic search was conducted on PubMed, EMBASE, MEDLINE, PsyclNFO, Web of Science, Clinical Trials.gov, and references in reviews from the earliest available date to September 2023. Randomized controlled trials (RCTs) were identified that evaluated the efficacy of bumetanide in improving overall core symptoms (OCS) of ASD. Therefore, nine studies with 1036 participants were included in the study. Bumetanide showed significant effects on OCS of ASD (WMD = － 1.91, p = 0.006), particularly in sub-domains including relation to inanimate objects, adaption to environment changes, auditory response, near sensory responses, anxiety and hyperactivity. Moderating analysis indicated that a significant effect size of bumetanide on OCS of ASD was observed in specific subgroup, including 3–6 years old (WMD = －1.08, p = 0.008), the tablet (WMD = －2.80, p = 0.003), 3-month intervention (WMD = －2.54, p = 0.003), and the single-center studies (WMD = －2.80, p = 0.003). Bumetanide has a large and significant impact on the OCS of ASD. Given the limited number and quality of included RCTs, future research should prioritize conducting large-scale trials focusing on sub-parameters or specific clinical features to comprehensively evaluate the efficacy of bumetanide in subpopulations of children with ASD. •Bumetanide showed significant and large effects on the overall core symptoms of ASD.•Bumetanide’s efficacy on ASD is influenced by subjects’ age, dosage form, duration.•Results of RCTs on bumetanide in ASD are moderated by study designs, measurement tools.