Probiotics in epilepsy treatment: a meta-analysis of efficacy and safety

Background Interest in probiotics for managing neurological disorders has grown recently. Methods Eight public databases were searched for relevant randomized controlled trials (RCTs) published until May 31, 2025. Eligible studies were screened based on predefined inclusion and exclusion criteria. Basic characteristics and outcome data were extracted from the included studies, and the risk of bias was assessed. Meta-analyses, sensitivity analyses, and publication bias evaluations were performed using RevMan 5.3 software. The quality of evidence was rated using GRADEpro GDT. Results Six RCTs involving 656 patients were included. Compared with the AED group, the probiotic combination group had significantly higher clinical efficacy rate (relative risk [RR] 1.16, 95% confidence interval [CI] 1.09 to 1.24, p < 0.00001) and lower seizure frequency (mean difference [MD] −1.97, 95% CI −2.35 to −1.59, p < 0.00001), seizure duration (MD −0.56, 95% CI −0.74 to −0.38, p < 0.00001), electroencephalogram (EEG) θ band relative power (MD −1.89, 95% CI −2.51 to −1.26, p < 0.00001), serum diamine oxidase (DAO) levels (MD −1.26, 95% CI −1.61 to −0.90, p < 0.00001), and serum D-lactate (D-LAC) levels (MD −2.04, 95% CI −2.78 to −1.30, p < 0.00001). No statistically significant differences were observed between the two groups in terms of interictal epileptiform discharge (MD −1.82, 95% CI −4.04 to 0.41, p = 0.11), EEG δ band relative power (MD 0.13, 95% CI −0.23 to 0.48, p = 0.49), EEG α band relative power (MD −0.27, 95% CI −1.34 to 0.80, p = 0.62), and EEG β band relative power (MD 0.10, 95% CI −0.17 to 0.38, p = 0.47). No significant difference was observed in the total adverse reaction rate between two groups (RR 0.58, 95% CI 0.32 to 1.08, p = 0.08). Funnel plots revealed no publication bias for clinical efficacy rate, seizure frequency, seizure duration, EEG β band relative power, and serum DAO and D-LAC levels. Conclusion Probiotics improved seizure activity, relative power in certain EEG frequency bands, and intestinal barrier function without increasing the incidence of adverse reactions, supporting their feasibility as complementary treatment for epilepsy. Limited sample size and methodological necessitate further validation. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420251079172 , Identifier CRD420251079172.