Effects of synbiotic supplementation on oxidative stress biomarkers in adults: an updated GRADE-assessed systematic review and meta-analysis of randomized controlled trials Authors: Mohammad Vesal Bideshki, Maedeh Nojoumi, Mehrdad Behzadi, Nasir Arefinia, Hedyeh Askarpour, Zahra Sohrabi, Mohammad Esmailian, Mohammad Ali-Hassanzadeh Published: 2026 ABSTRACT BackgroundSynbiotics (combined probiotics + prebiotics) may ameliorate oxidative stress via gut microbiota–mediated mechanisms. However, previous syntheses were limited in biomarker scope and contaminated by retracted or compromised trials. This updated meta-analysis of randomized controlled trials (RCTs) aims to quantify the effects of synbiotic supplementation on a comprehensive panel of oxidative stress biomarkers in adults and to appraise the certainty of evidence.MethodsFollowing PRISMA guidance and a registered protocol, RCTs in adults comparing oral synbiotic supplementation (≥ 3 weeks) versus placebo were eligible. Primary outcomes were changes in malondialdehyde (MDA), total oxidant status (TOS), oxidative stress index (OSI), nitric oxide (NO), total antioxidant capacity (TAC), superoxide dismutase (SOD), glutathione peroxidase (GPx), and reduced glutathione (GSH). Data were pooled using a random-effect model (standardized mean differences, SMD; 95% CI). Heterogeneity (I²), sensitivity, subgroup and publication-bias analyses were performed. Evidence certainty was rated with GRADE; studies formally retracted or with serious concerns were excluded.ResultsEighteen RCTs were included. Synbiotic supplementation significantly reduced MDA (SMD: -0.87; 95% CI: -1.37, -0.37), TOS (SMD: -0.55; 95% CI: -1.03, -0.08) and OSI (SMD: -0.46; 95% CI: -0.78, -0.15). Antioxidant markers, including TAC (SMD: 0.25; 95% CI: 0.06, 0.43), SOD (SMD: 0.57; 95% CI: 0.13, 1.00), GSH (SMD: 0.70; 95% CI: 0.11, 1.28), and GPx (SMD: 1.40; 95% CI: 0.39, 2.40), were increased. GRADE ratings ranged from very low (MDA, GPx) to moderate (OSI).ConclusionSynbiotic supplementation may improve certain oxidative stress biomarkers, including TAC, GSH, and MDA. However, findings for NO and OSI should be interpreted with caution due to the limited number of available studies and the low to very low certainty of evidence for some outcomes according to GRADE criteria. Further high-quality, large-scale RCTs are warranted to confirm these effects.Systematic review registrationhttps://www.crd.york.ac.uk/PROSPERO/view/CRD420251103060.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12263-026-00809-9. Introduction Oxidative stress arises from excess reactive oxygen species (ROS) overwhelming antioxidant defenses, causing damage to lipids, proteins, and DNA [1, 2]. This imbalance manifests as reduced total antioxidant capacity (TAC) and serum thiol levels, contributing to the pathogenesis of cardiovascular diseases, type 2 diabetes, and cancers [3]. Key enzymatic defenses include superoxide dismutase (SOD), which converts superoxide to hydrogen peroxide and water [4, 5], and glutathione peroxidase (GPx), which reduces lipid hydroperoxides and H₂O₂ [6–8]. Glutathione (GSH) serves as a critical non-enzymatic redox buffer [6–8]. In normal physiological states, the body achieves redox homeostasis by equilibrating pro-oxidant and antioxidant forces [9]. During disease progression, however, endogenous compensatory pathways often become insufficient, necessitating the use of extrinsic agents to reestablish oxidative balance [3, 10]. In recent years, synbiotics, combinations of probiotics (beneficial live microorganisms) and prebiotics (their growth-promoting substrates), have gained attention as potential modulators of oxidative stress through gut microbiota remodeling [11, 12]. Preliminary evidence suggestd that they could possibly support immune regulation [13–15], enhance antioxidant defenses [16, 17], and influence metabolic disorders, cognitive function, and malignancies via microbiome changes [18–21]. Proposed mechanisms include metal ion chelation, production of bioactive metabolites, ROS scavenging, and lipid profile adjustments that could lower malondialdehyde (MDA) [4, 17, 22, 23]. Synbiotics also generate short-chain fatty acids (SCFAs), which appear to facilitate GSH regeneration, attenuate pro-inflammatory interleukin-6 (IL-6), boost anti-inflammatory interleukin-18 (IL-18), and modulate oxidative stress genes [24, 25]. Butyrate limits peroxidation of low-density lipoprotein, 8-isoprostanes, and the glutathione redox ratio [26]. Further, certain methyl ketones produced under synbiotic influence could confer anti-inflammatory effects [27]. Existing meta-analyses on probiotic or synbiotic interventions have typically examined limited biomarkers (e.g., TAC, GSH, MDA, Nitric oxide (NO)) in specific groups, such as diabetic or renal patients [9, 16, 17, 28–31]. No review has assessed the broader array of markers, including SOD, GPx, total oxidant status (TOS), and oxidative stress index (OSI) across healthy adults and multiple conditions. While several meta-analyses have investigated the effects of synbiotics on oxidative stress [16, 32–34], a critical concern remains regarding the quality of the primary studies included in previous syntheses. Notably, recent scrutiny has revealed that some trials included in prior reviews have since been retracted or flagged for data integrity concerns [35–44], which can introduce significant bias and compromise the validity of pooled effect sizes. To address this gap, the present study rigorously identified and excluded retracted publications and studies with serious concerns to strengthen the evidence base. Therefore, this systematic review and meta-analysis aims to provide a robust evaluation of synbiotic effects on a comprehensive panel of eight oxidative stress biomarkers, utilizing the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) framework to ensure the certainty of findings across healthy and clinical adult populations. Materials and methods This systematic review was performed following the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines [45] and is registered with PROSPERO under the registration number of CRD420251103060. Search strategy A thorough systematic literature search was performed across PubMed, SCOPUS, and Web of Science to identify randomized controlled trials that met the predefined inclusion criteria. The search included peer-reviewed, English-language studies published through December 2025. Adhering to PRISMA guidelines, the search strategy employed a combination of Medical Subject Headings (MeSH) terms and keywords, such as “synbiotic supplementation,” “probiotic AND prebiotic,” “oxidative stress markers,” “Total Oxidant Status,” “Malondialdehyde,” “Nitric Oxide,” “Total Antioxidant Capacity,” “Superoxide Dismutase,” “Glutathione Peroxidase,” and “gut microbiota.” Manual screening of reference lists from relevant review articles was also conducted. Additionally, a supplementary search was carried out on Google Scholar, reviewing the first 10 pages of results sorted by relevance. Study selection The studies were selected using the Population, Intervention, Comparison, Outcome (PICO) model (Table 1) with the following inclusion criteria: (1) randomized controlled trials (RCTs) employing either crossover or parallel designs; (2) synbiotic supplementation administered for at least 3 weeks; (3) intervention and placebo groups differing only by the synbiotic supplementation; (4) participants aged 18 years or older, encompassing both healthy and unhealthy populations; (5) reported outcomes of synbiotic supplementation, including baseline and post-intervention values with standard deviations (SDs), standard errors (SEs), or 95% confidence intervals (CIs) for both groups; (6) synbiotic supplement delivered orally; and (7) studies published in peer-reviewed journals. Table 1Population, intervention, comparison, outcome, study design (PICO) criteria for inclusion and exclusion of studiesParameterCriteriaParticipantAdultsInterventionSynbioticComparatorMatched control groupOutcomesMDA, TAC, TOS, SOD, GSH, NO, GPX, OSIStudy designControlled clinical trialAbbreviation: MDA: Malondialdehyde; TAC: Total antioxidant capacity; TOS: Total oxidant status; SOD: superoxide dismutase; GSH: Glutathione; NO: Nitric oxide; GPX: glutathione peroxidase; OSI: oxidative stress index Table 1Population, intervention, comparison, outcome, study design (PICO) criteria for inclusion and exclusion of studiesParameterCriteriaParticipantAdultsInterventionSynbioticComparatorMatched control groupOutcomesMDA, TAC, TOS, SOD, GSH, NO, GPX, OSIStudy designControlled clinical trialAbbreviation: MDA: Malondialdehyde; TAC: Total antioxidant capacity; TOS: Total oxidant status; SOD: superoxide dismutase; GSH: Glutathione; NO: Nitric oxide; GPX: glutathione peroxidase; OSI: oxidative stress index Population, intervention, comparison, outcome, study design (PICO) criteria for inclusion and exclusion of studies Abbreviation: MDA: Malondialdehyde; TAC: Total antioxidant capacity; TOS: Total oxidant status; SOD: superoxide dismutase; GSH: Glutathione; NO: Nitric oxide; GPX: glutathione peroxidase; OSI: oxidative stress index Studies were excluded based on the following criteria: (1) non-human studies, reviews, or observational research; (2) absence of baseline or post-intervention data for the defined outcome measures; (3) trials involving short-term or acute supplementation periods; (4) retracted articles; (5) publications not in English; (6) studies lacking details on probiotic strains or prebiotic components; and (7) studies where the specific impact of synbiotic supplementation on the outcome measures could not be assessed. Data extraction Two researchers (MV.B. and M.B.) independently conducted the selection of eligible trials, data extraction, and evaluation of bias risk. Disagreements were resolved through consultation with a third investigator (M.N.). From the included studies, data was collected regarding the following: name of the first author, title of publication, the publication year, study design, location of study, sample sizes for intervention and placebo groups, health status, sex distribution, mean age at baseline, participants’ body mass index (BMI), features of the synbiotic supplement (including type, daily dosage, prebiotic type and amount, probiotic strain, and colony-forming units), duration of supplementation, placebo composition, and details of any concurrent interventions. Quality assessment The quality of the included studies was assessed using the Cochrane quality assessment tool, which categorized studies as having low, high, or unclear risk based on the detected biases. The overall quality and certainty of evidence were evaluated using the GRADE framework. This approach employs rigorous and transparent criteria to rate evidence quality as High, Moderate, Low, or Very Low. Evidence certainty could be downgraded due to factors including (1) bias risk (2), inconsistency across studies (3), indirectness (4), imprecision, or (5) publication bias. Conversely, certainty could be upgraded in cases such as a strong effect size, evidence of a dose-response relationship, or when plausible confounders strengthen confidence in the effect estimate. The GRADE framework provides a systematic method for evaluating evidence, supporting the formulation of recommendations by weighing intervention benefits against potential risks with clarity and reliability [46]. Statistical analysis The effect of synbiotic supplementation on oxidative stress biomarkers was quantified using standardized mean differences (SMDs) with 95% confidence intervals (CIs). For each trial, the primary effect estimate was calculated as the between-group difference in change from baseline (post-intervention minus baseline) for antioxidant defense biomarkers (GPx, GSH, SOD, TAC) and oxidative damage biomarkers (MDA, NO, OSI, TOS). When change scores were not reported, they were derived as (post-intervention − baseline) within each group. If the standard deviation (SD) of the change was unavailable, it was imputed using: SD_change = √[(SD_baseline)² + (SD_post)² − (2 × R × SD_baseline × SD_post)], assuming a correlation coefficient (R) of 0.8 [47]. For studies reporting standard error of the mean (SEM), SD was calculated as SD = SEM × √n. Pooled effects were synthesized using a random-effects model and reported as SMDs (95% CIs). Statistical heterogeneity was assessed using Cochran’s Q test and the I² statistic (p < 0.10 and/or I² ≥ 50% indicating substantial heterogeneity) [48]. Prespecified subgroup analyses explored potential sources of heterogeneity according to BMI, health condition, and intervention duration. Publication bias was evaluated using funnel plots, Begg’s rank correlation, and Egger’s regression tests; where indicated, trim-and-fill and fail-safe N methods were applied [49]. Statistical analyses in this meta-analysis were performed using Stata version 17 (Stata Corp, Texas, USA). A P-value < 0.05 was indicative of statistical significance across all tests. Results Flow and characteristics of included studies As illustrated in Fig. 1, a total of 3,697 records were initially identified through database searching. After removal of duplicates (n = 752), 2,945 records remained for screening. Following title and abstract assessment, 2,906 records were excluded as they did not meet the inclusion criteria, primarily because they were not randomized controlled trials or were not directly relevant to the research question. Consequently, 43 articles were deemed potentially eligible and were retrieved for full-text evaluation. Of these, 25 studies were subsequently excluded for the following reasons: retracted articles or have some concerns (n = 10) [35–44], non-Randomized or uncontrolled clinical trial (n = 3) [50–52], no measurements of the factors (n = 7) [53–59], short-term synbiotic supplementation (n = 2) [60, 61], age under 18 years (n = 2) [62, 63], and measuring factors in breastmilk (n = 1) [64]. Fig. 1Flow diagram of systematic literature search Flow diagram of systematic literature search Following the inclusion criteria, 18 qualified RCTs were selected for the final analysis [23, 55, 59, 65–79]. According to Table 1, the included studies in this review met the PICOS criteria. Characteristics of the included studies The detailed characteristics of the 18 randomized controlled trials included in this meta-analysis are summarized in Table 2. These studies were published between 2017 [65] and 2025 [78, 79]. Geographically, the majority of the trials were conducted in Iran (13 studies) [23, 55, 59, 65, 67–73, 75, 76], while others were performed in Thailand (2 studies) [74, 77], Poland (1 study) [66], Egypt (1 study) [79], and China (1 study) [78]. All studies employed a randomized design, with the majority utilizing a double-blind, parallel-group structure. Table 2Characteristic of included studies in meta-analysisStudiesCountryStudy DesignParticipantSample size and SexSample sizeTrial Duration(Week)Means AgeMeans BMIInterventionFeatures of InterventionControl groupProbioticsPrebioticsIGCGIGCGIGCGType of colonyDose of colony (CFU)Type of prebiotisDosage(mg)Ekhlasi et al. (2017 A)IranRn/Db/PaNAFLD patients30B15158NRNR27.2827.84Synbiotic capsuleLactobacillus casei, Lactobacillus rhamnosus, Stretococcus thermophilus, Bifidobacterium breve, Lactobacillus acidophilus, Bifidobacterium longum, Lactobacillus bulgaricus7 × 4 × 10 8FOSNRCorn starchEkhlasi et al. (2017 B)IranRn/Db/PaNAFLD patients30B15158NRNR28.0528.77Synbiotic capsule+Alpha-tocopherol(400 IU)Lactobacillus casei, Lactobacillus rhamnosus, Stretococcus thermophilus, Bifidobacterium breve, Lactobacillus acidophilus, Bifidobacterium longum and Lactobacillus bulgaricus7 × 4 × 10 8FOSNRCorn starch + Alpha-tocopherol (400 IU)Kleniewska et al. (2017)PolandRn/PaHealthy volunteers32B16167NRNRNRNRSynbiotic capsule Lactobacillus casei 4 × 10 8Inulin400PlaceboBakhshimoghaddam et al. (2018)IranRn/PaNAFLD patients68B34342438.839.930.531.3Synbiotic yougurt Bifidobacterium animalis subsp. lactis (BB-12) 1 × 10 8Inulin1500Conventional yougurtNabhani et al. (2018)IranRn/Db/PaGestational diabetes mellitus patients90F4545629.430.326.428.2Synbiotic capsuleLactobacillus acidophilus, Lactobacillus plantarum, Lactobacillus fermentum, Lactobacillus Gasseri9.2 × 10 9FOS38.5Lactose (300 mg), magnesium stearate, talc, colloidal silicon dioxide (each of them 5.5 mg), flavorings and sweetenersGhafouri et al. (2019 A)IranRn/Db/PaT2DM patients50B2525854.9254.626.3927.04Synbiotic bread Bacilluscoagulans 1 × 10 8Inulin10,000Bread containing beta-glucan (3 g)Ghafouri et al. (2019 B)IranRn/Db/PaT2DM patients50B2525853.885526.8326.33Synbiotic+lactic acid bread Bacilluscoagulans 1 × 10 8Inulin10,000Lactic acid (4 g)Navaei et al. (2019)IranRn/Db/PaSurvivors of Breast Cancer88F44441053/852/3930/9330/88Synbiotic capsuleLactobacillus casei, Lactobacillus acidophilus, Lactobacillus rhamnosus, Lactobacillus bulgaricus, Bifidobacterium breve, Bifidobacterium longum, and Streptococcus thermophilus9 × 10 9FOS38.5placebo + low-calorie dietSoleimani et al. (2019)IranRn/Db/PaDiabetic patients undergoing HD60B30301262.862.826.426.9Synbiotic capsuleLactobacillus acidophilus strain T16 (IBRCM10785), L. casei strain T2 (IBRC-M10783), and Bifidobacterium bifidum strain T1 (IBRC-M10771)6 × 10 9Inulin800Corn starchKooshki et al. (2019)IranRn/Db/PaHD patients46B2323862.9262.8323.4523.6Synbiotic capsule Lactobacillus coagulans NRFOSNRPlaceboBazyar et al. (2020)IranRn/Db/PaT2DM patients47B2324848.650.12425.5Synbiotic capsuleContained 7 viable and freeze-dried strains:Lactobacillus acidophilus, Lactobacillus casei, Lactobacillus rhamnosus, Lactobacillus bulgaricus, Bifidobacterium breve, Bifidobacterium longum and Streptococcus thermophilus23 × 10 9FOS100PlaceboChaiyasut et al. (2021)ThailandRn/Db/PaObese adults72B36361254.7858.9428.9730.01Synbiotic sachetsLactobacillus paracasei, Bifidobacterium longum, Bifidobacterium breve,5 × 10 10Inulin and FOS10,000Corn starchGhavami et al. (2021)IranRn/Db/PaMigraine patients80F40401238.0537.6227.7427.38Synbiotic capsuleLactobacillus casei, Lactobacillus acidophilus, Lactobacillus rhamnosus, Lactobacillus helveticus, Lactobacillus bulgaricus, Lactobacillus plantarum, Lactobacillus gasseri, Bifidobacterium breve, Bifidobacterium longum, Bifidobacterium lactis, Bifidobacterium bifidum, and Streptococcus thermophilus,2 × 10 9FOS42StarchNabhani et al. (2022)IranRn/Db/PaPregnant women in weeks 24–28 of pregnancy with Gestational diabetes mellitus90F4545629.430.326.428.2Synbiotic capsuleLactobacillus acidophilus, Lactobacillus plantarum, Lactobacillus fermentum and Lactobacillus gasseri9.2 × 10 10FOS - lactose338.5PlaceboFallah et al. (2023)IranRn/Db/PaObese patients46F2323838.393832.5932.22Synbiotic + L-CarnitineLactobacillus casei PXN 37, Lactobacillus rhamnosus PXN 54, Lactobacillus acidophilus PXN 35, Lactobacillus bulgaricus PXN 39, Bifidobacterium breve PXN 25, Bifidobacterium longum PXN 30 and Streptococcus thermophilus PXN 661 × 10 8FOS175Same amounts of L-carnitine+ MaltodextrinMehrabani et al. (2023)IranRn/Db/PaPatients with Parkinson’s disease80B40401268.269.05NRNRSynbiotic sachetLactobacillus acidophilus (LAA-5), Lactobacillus rhamnosus (LAR-7), Lactobacillus plantarum (LAP-10), Bifidobacterium longum (BIA-8), Streptococcus thermophilus5 × 10 9Inulin4000MaltodextrinOraphruek et al. (2023)ThailandRn/Db/PaOverweight and Obese patients63B32311232/3533/6927/4727Synbiotic sachetLactobacillus rhamnosus LR3, Lactobacillus gasseri BNR17, Lactobacillus salivarius LS1, Bifidobacterium lactis BL2, Bifidobacterium longum BG3, Bifidobacterium breve BR2, Bifidobacterium infantis BT37 × 10 9FOS2000MaltodextrinZolghadrpour et al. (2024)IranRn/Db/PaMetabolic syndrome patients41B22191245/5543/7929/9728/98Synbiotic yogurtLactobacillus plantarum, Lactobacillus pentosus2 × 10 8NRNRRegular yogurtRamadan et al. (2025)EgyptRn/PaPatients with Parkinson’s disease66B33331256/0655/7NRNRSynbiotic sachet Lactobacillus acidophilus 1 × 10 10Inulin3000PlaceboZhang et al. (2025)ChinaRn/Db/PaT2DMpatients80B404012586124/725Synbiotic sachet Bifidobacterium animalis subsp. Lactis 5 × 10 10Galactooligosaccharide2500PlaceboAbbreviations: IG, intervention group; CG, control group; Db, double-blinded; Sb, single-blinded; Pa, Parallel; Co, Cross-Over; Rn, randomized; NR, not reported; F, Female; B, Both gender; NR, not reported; CFU, Colony Forming Unit; FOS, Fructooligosaccharides; NAFLD, Non-alcoholic fatty liver disease; T2DM, Type 2 diabetes mellitus; HD, Hemodialysis Table 2Characteristic of included studies in meta-analysisStudiesCountryStudy DesignParticipantSample size and SexSample sizeTrial Duration(Week)Means AgeMeans BMIInterventionFeatures of InterventionControl groupProbioticsPrebioticsIGCGIGCGIGCGType of colonyDose of colony (CFU)Type of prebiotisDosage(mg)Ekhlasi et al. (2017 A)IranRn/Db/PaNAFLD patients30B15158NRNR27.2827.84Synbiotic capsuleLactobacillus casei, Lactobacillus rhamnosus, Stretococcus thermophilus, Bifidobacterium breve, Lactobacillus acidophilus, Bifidobacterium longum, Lactobacillus bulgaricus7 × 4 × 10 8FOSNRCorn starchEkhlasi et al. (2017 B)IranRn/Db/PaNAFLD patients30B15158NRNR28.0528.77Synbiotic capsule+Alpha-tocopherol(400 IU)Lactobacillus casei, Lactobacillus rhamnosus, Stretococcus thermophilus, Bifidobacterium breve, Lactobacillus acidophilus, Bifidobacterium longum and Lactobacillus bulgaricus7 × 4 × 10 8FOSNRCorn starch + Alpha-tocopherol (400 IU)Kleniewska et al. (2017)PolandRn/PaHealthy volunteers32B16167NRNRNRNRSynbiotic capsule Lactobacillus casei 4 × 10 8Inulin400PlaceboBakhshimoghaddam et al. (2018)IranRn/PaNAFLD patients68B34342438.839.930.531.3Synbiotic yougurt Bifidobacterium animalis subsp. lactis (BB-12) 1 × 10 8Inulin1500Conventional yougurtNabhani et al. (2018)IranRn/Db/PaGestational diabetes mellitus patients90F4545629.430.326.428.2Synbiotic capsuleLactobacillus acidophilus, Lactobacillus plantarum, Lactobacillus fermentum, Lactobacillus Gasseri9.2 × 10 9FOS38.5Lactose (300 mg), magnesium stearate, talc, colloidal silicon dioxide (each of them 5.5 mg), flavorings and sweetenersGhafouri et al. (2019 A)IranRn/Db/PaT2DM patients50B2525854.9254.626.3927.04Synbiotic bread Bacilluscoagulans 1 × 10 8Inulin10,000Bread containing beta-glucan (3 g)Ghafouri et al. (2019 B)IranRn/Db/PaT2DM patients50B2525853.885526.8326.33Synbiotic+lactic acid bread Bacilluscoagulans 1 × 10 8Inulin10,000Lactic acid (4 g)Navaei et al. (2019)IranRn/Db/PaSurvivors of Breast Cancer88F44441053/852/3930/9330/88Synbiotic capsuleLactobacillus casei, Lactobacillus acidophilus, Lactobacillus rhamnosus, Lactobacillus bulgaricus, Bifidobacterium breve, Bifidobacterium longum, and Streptococcus thermophilus9 × 10 9FOS38.5placebo + low-calorie dietSoleimani et al. (2019)IranRn/Db/PaDiabetic patients undergoing HD60B30301262.862.826.426.9Synbiotic capsuleLactobacillus acidophilus strain T16 (IBRCM10785), L. casei strain T2 (IBRC-M10783), and Bifidobacterium bifidum strain T1 (IBRC-M10771)6 × 10 9Inulin800Corn starchKooshki et al. (2019)IranRn/Db/PaHD patients46B2323862.9262.8323.4523.6Synbiotic capsule Lactobacillus coagulans NRFOSNRPlaceboBazyar et al. (2020)IranRn/Db/PaT2DM patients47B2324848.650.12425.5Synbiotic capsuleContained 7 viable and freeze-dried strains:Lactobacillus acidophilus, Lactobacillus casei, Lactobacillus rhamnosus, Lactobacillus bulgaricus, Bifidobacterium breve, Bifidobacterium longum and Streptococcus thermophilus23 × 10 9FOS100PlaceboChaiyasut et al. (2021)ThailandRn/Db/PaObese adults72B36361254.7858.9428.9730.01Synbiotic sachetsLactobacillus paracasei, Bifidobacterium longum, Bifidobacterium breve,5 × 10 10Inulin and FOS10,000Corn starchGhavami et al. (2021)IranRn/Db/PaMigraine patients80F40401238.0537.6227.7427.38Synbiotic capsuleLactobacillus casei, Lactobacillus acidophilus, Lactobacillus rhamnosus, Lactobacillus helveticus, Lactobacillus bulgaricus, Lactobacillus plantarum, Lactobacillus gasseri, Bifidobacterium breve, Bifidobacterium longum, Bifidobacterium lactis, Bifidobacterium bifidum, and Streptococcus thermophilus,2 × 10 9FOS42StarchNabhani et al. (2022)IranRn/Db/PaPregnant women in weeks 24–28 of pregnancy with Gestational diabetes mellitus90F4545629.430.326.428.2Synbiotic capsuleLactobacillus acidophilus, Lactobacillus plantarum, Lactobacillus fermentum and Lactobacillus gasseri9.2 × 10 10FOS - lactose338.5PlaceboFallah et al. (2023)IranRn/Db/PaObese patients46F2323838.393832.5932.22Synbiotic + L-CarnitineLactobacillus casei PXN 37, Lactobacillus rhamnosus PXN 54, Lactobacillus acidophilus PXN 35, Lactobacillus bulgaricus PXN 39, Bifidobacterium breve PXN 25, Bifidobacterium longum PXN 30 and Streptococcus thermophilus PXN 661 × 10 8FOS175Same amounts of L-carnitine+ MaltodextrinMehrabani et al. (2023)IranRn/Db/PaPatients with Parkinson’s disease80B40401268.269.05NRNRSynbiotic sachetLactobacillus acidophilus (LAA-5), Lactobacillus rhamnosus (LAR-7), Lactobacillus plantarum (LAP-10), Bifidobacterium longum (BIA-8), Streptococcus thermophilus5 × 10 9Inulin4000MaltodextrinOraphruek et al. (2023)ThailandRn/Db/PaOverweight and Obese patients63B32311232/3533/6927/4727Synbiotic sachetLactobacillus rhamnosus LR3, Lactobacillus gasseri BNR17, Lactobacillus salivarius LS1, Bifidobacterium lactis BL2, Bifidobacterium longum BG3, Bifidobacterium breve BR2, Bifidobacterium infantis BT37 × 10 9FOS2000MaltodextrinZolghadrpour et al. (2024)IranRn/Db/PaMetabolic syndrome patients41B22191245/5543/7929/9728/98Synbiotic yogurtLactobacillus plantarum, Lactobacillus pentosus2 × 10 8NRNRRegular yogurtRamadan et al. (2025)EgyptRn/PaPatients with Parkinson’s disease66B33331256/0655/7NRNRSynbiotic sachet Lactobacillus acidophilus 1 × 10 10Inulin3000PlaceboZhang et al. (2025)ChinaRn/Db/PaT2DMpatients80B404012586124/725Synbiotic sachet Bifidobacterium animalis subsp. Lactis 5 × 10 10Galactooligosaccharide2500PlaceboAbbreviations: IG, intervention group; CG, control group; Db, double-blinded; Sb, single-blinded; Pa, Parallel; Co, Cross-Over; Rn, randomized; NR, not reported; F, Female; B, Both gender; NR, not reported; CFU, Colony Forming Unit; FOS, Fructooligosaccharides; NAFLD, Non-alcoholic fatty liver disease; T2DM, Type 2 diabetes mellitus; HD, Hemodialysis Characteristic of included studies in meta-analysis 30 B 30 B 32 B 68 B 90 F 50 B 50 B 88 F 60 B 46 B 47 B Contained 7 viable and freeze-dried strains: Lactobacillus acidophilus, Lactobacillus casei, Lactobacillus rhamnosus, Lactobacillus bulgaricus, Bifidobacterium breve, Bifidobacterium longum and Streptococcus thermophilus 72 B 80 F 90 F 46 F 80 B 63 B 41 B 66 B T2DM patients 80 B Abbreviations: IG, intervention group; CG, control group; Db, double-blinded; Sb, single-blinded; Pa, Parallel; Co, Cross-Over; Rn, randomized; NR, not reported; F, Female; B, Both gender; NR, not reported; CFU, Colony Forming Unit; FOS, Fructooligosaccharides; NAFLD, Non-alcoholic fatty liver disease; T2DM, Type 2 diabetes mellitus; HD, Hemodialysis The study populations were heterogeneous, encompassing both healthy individuals and patients with various clinical conditions. These conditions included Type 2 Diabetes Mellitus (T2DM) [69, 73, 78], Non-alcoholic Fatty Liver Disease (NAFLD) [66, 67], Gestational Diabetes Mellitus (GDM) [68, 76], obesity and overweight [23, 74, 77], Parkinson’s disease [55, 79], metabolic syndrome [59], migraine headache [75], and patients undergoing hemodialysis [71, 72]. Additionally, one study involved survivors of breast cancer [71]. The participants included both genders in most studies, while some trials recruited exclusively female participants [23, 68, 71, 75, 76]. The duration of synbiotic supplementation ranged from 6 [68, 76] to 24 [67] weeks, with the most common duration being 8 to 12 weeks. The daily dosage of probiotic strains ranged from 1 × 10 8 [23, 67, 69] to 9.2 × 10 10 [76] colony forming unit (CFU). The probiotic components primarily consisted of strains from the Lactobacillus, Bifidobacterium, and Streptococcus genera, as well as Bacillus coagulans in some interventions. Prebiotic fibers used in the symbiotic formulations included inulin, fructooligosaccharides, and galactooligosaccharides, with varying dosages. Data quality Table 3 presents the results of the Cochrane risk of bias assessment for the included studies. Based on the evaluation, the studies were categorized into three quality levels: Good, Fair, and Poor, according to the number of domains with a low risk of bias. The majority of the studies (10 studies) were classified as having good quality [23, 55, 65, 69, 71, 73–76, 78]. These studies reflect a robust methodological framework, enhancing the reliability of their findings. A smaller subset of 4 studies was identified as fair quality [68, 70, 72, 77]. Lastly, 4 studies were determined to have poor quality [59, 66, 67, 79]. These studies present a higher likelihood of methodological limitations, necessitating careful consideration when synthesizing their results. Table 3Risk of bias assessmentStudyRandom sequence generationAllocation concealmentSelective reportingOther sources of biasBlinding (participants and personnel)Blinding (outcome assessment)Incomplete outcome dataGeneral QualityEkhlasi et al. (2017)LLLLLLLGoodKleniewska et al. (2017)UULUHHUPoorBakhshimoghaddam et al. (2018)LHLLHHLPoorNabhani et al. (2018)LLLLLLHFairGhafouri et al. (2019)LLLULLHGoodNavaei et al. (2019)LLLLLLHFairSoleimani et al. (2019)LLLULLLGoodKooshki et al. (2019)LLUUUUUFairBazyar et al. (2020)LLLLLLLGoodChaiyasut et al. (2021)LULULUUGoodGhavami et al. (2021)LLLULLHGoodNabhani et al. (2022)LLLLLLLGoodFallah et al. (2023)LLLLLLLGoodMehrabani et al. (2023)LLLLLLHGoodOraphruek et al. (2023)LLLULLUFairZolghadrpour et al. (2024)LULUUUHPoorRamadan et al. (2025)LLLLHHLPoorZhang et al. (2025)LLLLLLLGoodAbbreviations: L: low risk of bias; H: high risk of bias; U: unclear risk of bias Table 3Risk of bias assessmentStudyRandom sequence generationAllocation concealmentSelective reportingOther sources of biasBlinding (participants and personnel)Blinding (outcome assessment)Incomplete outcome dataGeneral QualityEkhlasi et al. (2017)LLLLLLLGoodKleniewska et al. (2017)UULUHHUPoorBakhshimoghaddam et al. (2018)LHLLHHLPoorNabhani et al. (2018)LLLLLLHFairGhafouri et al. (2019)LLLULLHGoodNavaei et al. (2019)LLLLLLHFairSoleimani et al. (2019)LLLULLLGoodKooshki et al. (2019)LLUUUUUFairBazyar et al. (2020)LLLLLLLGoodChaiyasut et al. (2021)LULULUUGoodGhavami et al. (2021)LLLULLHGoodNabhani et al. (2022)LLLLLLLGoodFallah et al. (2023)LLLLLLLGoodMehrabani et al. (2023)LLLLLLHGoodOraphruek et al. (2023)LLLULLUFairZolghadrpour et al. (2024)LULUUUHPoorRamadan et al. (2025)LLLLHHLPoorZhang et al. (2025)LLLLLLLGoodAbbreviations: L: low risk of bias; H: high risk of bias; U: unclear risk of bias Risk of bias assessment Abbreviations: L: low risk of bias; H: high risk of bias; U: unclear risk of bias The GRADE framework was employed to evaluate the quality of evidence concerning the impact of synbiotic supplementation on oxidative stress biomarkers (Table S1). The findings revealed substantial variability in the certainty of evidence across different outcomes: OSI was graded as moderate (⊕⊕⊕◯), suggesting reasonable reliability. TAC, GSH, SOD, TOS, and NO demonstrated low certainty (⊕⊕◯◯), with precision and moderate heterogeneity being the key limiting factors. MDA and GPX were classified with very low certainty (⊕◯◯◯), reflecting significant concerns about inconsistency and the potential presence of publication bias. Meta-analysis Effect of synbiotic supplementation on MDA concentration A total of 13 studies, encompassing 14 treatment arms and including 791 participants (synbiotic = 397, and control = 394), assessed the effects of synbiotic supplementation on MDA levels. According to the results of the random-effect model, supplementation with synbiotics significantly reduced MDA levels (SMD: -0.87, 95% CI: -1.37 to -0.37, P < 0.01). However, considerable heterogeneity was observed among the studies (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{I}}^{2}\:$$\end{document}= 90.6%, P < 0.001) (Fig. 2A). Subgroup analysis of synbiotic supplementation on MDA is shown in Table 4. Based on subgroup analysis, significant reductions in MDA were observed in studies with fewer than 60 participants (P = 0.003) as well as those with 60 or more participants (P = 0.048). Moreover, reductions were also evident in trials involving participants aged ≥ 50 years (P = 0.037), studies with a duration of < 12 weeks (P < 0.001), and studies rated as good quality (P = 0.001). Fig. 2Forest plot for the effect of Synbiotic supplementation on A: Malondialdehyde, B: Total antioxidant capacity, C: Total oxidant status, D: superoxide dismutase, E: Glutathione, F: Nitric oxide, G: glutathione peroxidase, H: oxidative stress index, in adults. Data expressed as mean differences between intervention and control groups Forest plot for the effect of Synbiotic supplementation on A: Malondialdehyde, B: Total antioxidant capacity, C: Total oxidant status, D: superoxide dismutase, E: Glutathione, F: Nitric oxide, G: glutathione peroxidase, H: oxidative stress index, in adults. Data expressed as mean differences between intervention and control groups Table 4Pooled estimates of the subgroup analysesESSMD (95%CI)P-valueheterogeneityP heterogeneityI2P between sub-groups Subgroup analyses of synbiotic supplementation on TAC Overall effect 120.28 (0.07, 0.48) 0.008 0.028 49.1% Number of participants < 6050.38 (0.01, 0.74) 0.044 0.1048.7%0.502 ≥ 6070.22 (-0.03, 0.48)0.0820.0453.8% Mean age < 5060.37 (0.08, 0.67) 0.013 0.0849.1%0.366 ≥ 5060.18 (-0.1, 0.47)0.210.0652% Trial duration (week) < 1260.18 (-0.02, 0.38)0.0830.830%0.349 ≥ 1260.39 (0.00, 0.78) 0.049 0.002 73.1% Type of intervention Capsules or Sachets80.21 (-0.01, 0.44)0.0610.0943.8%0.416 Enriched food40.42 (-0.03, 0.87)0.0650.05161.4% Study quality Good40.45 (0.16, 0.74) 0.002 0.3115.6%0.107 Fair50.09 (-0.12, 0.29)0.410.700% Poor30.45 (-0.27, 1.17)0.427 0.004 81.8% Subgroup analyses of synbiotic supplementation on GSH Overall effect 70.7 (0.11, 1.28) 0.021 < 0.001 88% Number of participants < 6030.60 (-0.27, 1.46)0.176 0.004 82.3%0.791 ≥ 6040.76 (-0.11, 1.63)0.086 < 0.001 92.1% Trial duration (week) < 1230.60 (-0.27, 1.46)0.176 0.004 82.3%0.791 ≥ 1240.76 (-0.11, 1.63)0.086 < 0.001 92.1% Type of intervention Capsules or Sachets50.92 (0.14, 1.71) 0.021 < 0.001 90.7%0.081 Enriched food20.14 (-0.25, 0.54)0.4750.4660% Study quality Good31.06 (0.09, 2.04) 0.033 < 0.001 91.1%0.202 Fair20.37 (-0.05, 0.78)0.4750.4660% Poor20.73 (-1.01, 2.47)0.41 < 0.001 92.8% Subgroup analyses of synbiotic supplementation on MDA Overall effect 14-0.87 (-1.37, -0.37) 0.001 < 0.001 90.6% Number of participants < 607-1.04 (-1.74, -0.34) 0.003 < 0.001 85.6%0.508 ≥ 607-0.71 (-1.41, -0.01) 0.048 < 0.001 93% Mean age < 505-1.02 (-2.29, 0.24)0.113 < 0.001 95.5%0.538 ≥ 506-0.59 (-1.14, -0.04) 0.037 < 0.001 86.5% Trial duration (week) < 128-1.01 (-1.54, -0.48) < 0.001 < 0.001 83.5%0.545 ≥ 126-0.66 (-1.62, 0.27)0.163 < 0.001 94,6% Study quality Good7-0.82 (-1.31, -0.33) 0.001 < 0.001 80.4%0.658 Fair2-0.41 (-1.65, 0.83)0.204 < 0.001 97.7% Poor5-0.53 (-1.51, 0.46)0.296 < 0.001 92.7% Subgroup analyses of synbiotic supplementation on SOD Overall effect 70.567 (0.13, 1) 0.011 < 0.001 76.4% Number of participants < 6050.65 (0.10, 1.20) 0.020 < 0.01 74.1%0.637 ≥ 6020.39 (-0.56, 1.33)0.421 < 0.01 88.6% Mean age < 5021.16 (0.59, 1.74) < 0.001 0.2036.5%0.075 ≥ 5040.48 (-0.01, 0.97)0.054 0.01 73.1% Trial duration (week) < 1250.56 (0.14, 0.97) 0.008 0.02 62.9%0.89 ≥ 1220.67 (-0.88, 2.22)0.396 < 0.001 92.7% Type of intervention Capsules or Sachets40.36 (-0.25, 0.96)0.247 < 0.01 80.3%0.273 Enriched food30.86 (0.20, 1.51) 0.011 0.03 71.4% Study quality Good10.90 (0.29, 1.50) 0.003 < 0.001 0%0.652 Fair30.69 (0.33, 1.06)0.5020.2232.4% Poor30.35 (-0.67, 1.37) < 0.001 < 0.001 88.0%Abbreviations: CI, confidence interval; SMD, standard mean differences; ES, Effect size; TAC, total antioxidant capacity; GSH, total Glutathione; MDA, malondialdehyde; SOD, superoxide dismutase Table 4Pooled estimates of the subgroup analysesESSMD (95%CI)P-valueheterogeneityP heterogeneityI2P between sub-groups Subgroup analyses of synbiotic supplementation on TAC Overall effect 120.28 (0.07, 0.48) 0.008 0.028 49.1% Number of participants < 6050.38 (0.01, 0.74) 0.044 0.1048.7%0.502 ≥ 6070.22 (-0.03, 0.48)0.0820.0453.8% Mean age < 5060.37 (0.08, 0.67) 0.013 0.0849.1%0.366 ≥ 5060.18 (-0.1, 0.47)0.210.0652% Trial duration (week) < 1260.18 (-0.02, 0.38)0.0830.830%0.349 ≥ 1260.39 (0.00, 0.78) 0.049 0.002 73.1% Type of intervention Capsules or Sachets80.21 (-0.01, 0.44)0.0610.0943.8%0.416 Enriched food40.42 (-0.03, 0.87)0.0650.05161.4% Study quality Good40.45 (0.16, 0.74) 0.002 0.3115.6%0.107 Fair50.09 (-0.12, 0.29)0.410.700% Poor30.45 (-0.27, 1.17)0.427 0.004 81.8% Subgroup analyses of synbiotic supplementation on GSH Overall effect 70.7 (0.11, 1.28) 0.021 < 0.001 88% Number of participants < 6030.60 (-0.27, 1.46)0.176 0.004 82.3%0.791 ≥ 6040.76 (-0.11, 1.63)0.086 < 0.001 92.1% Trial duration (week) < 1230.60 (-0.27, 1.46)0.176 0.004 82.3%0.791 ≥ 1240.76 (-0.11, 1.63)0.086 < 0.001 92.1% Type of intervention Capsules or Sachets50.92 (0.14, 1.71) 0.021 < 0.001 90.7%0.081 Enriched food20.14 (-0.25, 0.54)0.4750.4660% Study quality Good31.06 (0.09, 2.04) 0.033 < 0.001 91.1%0.202 Fair20.37 (-0.05, 0.78)0.4750.4660% Poor20.73 (-1.01, 2.47)0.41 < 0.001 92.8% Subgroup analyses of synbiotic supplementation on MDA Overall effect 14-0.87 (-1.37, -0.37) 0.001 < 0.001 90.6% Number of participants < 607-1.04 (-1.74, -0.34) 0.003 < 0.001 85.6%0.508 ≥ 607-0.71 (-1.41, -0.01) 0.048 < 0.001 93% Mean age < 505-1.02 (-2.29, 0.24)0.113 < 0.001 95.5%0.538 ≥ 506-0.59 (-1.14, -0.04) 0.037 < 0.001 86.5% Trial duration (week) < 128-1.01 (-1.54, -0.48) < 0.001 < 0.001 83.5%0.545 ≥ 126-0.66 (-1.62, 0.27)0.163 < 0.001 94,6% Study quality Good7-0.82 (-1.31, -0.33) 0.001 < 0.001 80.4%0.658 Fair2-0.41 (-1.65, 0.83)0.204 < 0.001 97.7% Poor5-0.53 (-1.51, 0.46)0.296 < 0.001 92.7% Subgroup analyses of synbiotic supplementation on SOD Overall effect 70.567 (0.13, 1) 0.011 < 0.001 76.4% Number of participants < 6050.65 (0.10, 1.20) 0.020 < 0.01 74.1%0.637 ≥ 6020.39 (-0.56, 1.33)0.421 < 0.01 88.6% Mean age < 5021.16 (0.59, 1.74) < 0.001 0.2036.5%0.075 ≥ 5040.48 (-0.01, 0.97)0.054 0.01 73.1% Trial duration (week) < 1250.56 (0.14, 0.97) 0.008 0.02 62.9%0.89 ≥ 1220.67 (-0.88, 2.22)0.396 < 0.001 92.7% Type of intervention Capsules or Sachets40.36 (-0.25, 0.96)0.247 < 0.01 80.3%0.273 Enriched food30.86 (0.20, 1.51) 0.011 0.03 71.4% Study quality Good10.90 (0.29, 1.50) 0.003 < 0.001 0%0.652 Fair30.69 (0.33, 1.06)0.5020.2232.4% Poor30.35 (-0.67, 1.37) < 0.001 < 0.001 88.0%Abbreviations: CI, confidence interval; SMD, standard mean differences; ES, Effect size; TAC, total antioxidant capacity; GSH, total Glutathione; MDA, malondialdehyde; SOD, superoxide dismutase Pooled estimates of the subgroup analyses Abbreviations: CI, confidence interval; SMD, standard mean differences; ES, Effect size; TAC, total antioxidant capacity; GSH, total Glutathione; MDA, malondialdehyde; SOD, superoxide dismutase Effect of synbiotic supplementation on TAC concentration Across 11 studies involving 12 treatment arms and a total of 772 individuals (synbiotic = 387, and control = 385), changes in TAC levels following synbiotic supplementation were evaluated. Results from the random-effect model showed that synbiotic supplementation had a modest but significant increase in TAC levels (SMD: 0.25, 95% CI: 0.06 to 0.43, P < 0.01) with moderate heterogeneity among studies (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{I}}^{2\:}$$\end{document}= 47.6%, P = 0.02) (Fig. 2B). As presented in Table 4, subgroup analysis revealed a significant increasing in TAC levels were observed in studies with fewer than 60 participants (P = 0.044). Furthermore, individuals aged < 50 years (P = 0.01), in trials with durations of ≥ 12 weeks (P = 0.04), and studies of good quality (P = 0.002). Effect of synbiotic supplementation on TOS concentration Changes in TOS levels after synbiotic supplementation were evaluated in four studies comprising four treatment arms and a total of 269 individuals (synbiotic = 136, and control = 133). Pooled analysis using a random-effect model demonstrated a statistically significant decrease in TOS levels (SMD: -0.55, 95% CI: -1.03 to -0.08, P = 0.02). Significant heterogeneity was observed among the studies (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{I}}^{2\:}$$\end{document}= 72.4%, P = 0.01) (Fig. 2C). Subgroup analyses were not conducted for TOS due to the limited number of included studiesfor this variable. Effect of synbiotic supplementation on SOD concentration The effect of synbiotic supplementation on SOD levels was assessed across six studies, comprising seven treatment arms and 380 participants (synbiotic = 191, controls = 189). Based on pooled results from the random-effects model, synbiotic supplementation was associated with a statistically significant increase in SOD levels (SMD = 0.57, 95% CI: 0.13 to 1.00; P = 0.01). Considerable heterogeneity was observed across the included studies (I² = 76.4%, P < 0.001) (Fig. 2D). As detailed in Table 4, significant increasing in SOD levels were observed in trials enrolling participants younger than 60 years (P = 0.02) and under 50 years (P < 0.001), in studies with durations of < 12 weeks (P = 0.008), in both good and poor quality studies (P = 0.003, P < 0.001, respectivly), and in those employing enriched-food interventions (P = 0.01). Effect of synbiotic supplementation on GSH concentration GSH data were collected from six studies, which included seven treatment arms and a total of 424 individuals (synbiotic = 212, and control = 212). The random-effects meta-analysis showed that synbiotic supplementation was associated with a significant increase in GSH concentrations (SMD = 0.70, 95% CI: 0.11 to 1.28; P = 0.01). Despite the significant pooled effect, considerable between-study heterogeneity was identified (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{I}}^{2}$$\end{document} = 88.0%, P < 0.001) (Fig. 2E). Considering analysis of subgroup, significant increases were noted in those employing Capsules or Sachets interventions (P = 0.02), and in good quality studies (P = 0.03) (Table 4). Effect of synbiotic supplementation on NO concentration NO data were collected from three studies, which included four treatment arms and a total of 200 participants (synbiotic = 100, and control = 100). Using a random-effect model, the meta-analysis determined that NO levels were not significantly reduced following the consumption of synbiotics (SMD: -0.05; 95% CI: -0.48 to 0.38; P = 0.82). However, considerable heterogeneity was not detected among the studies (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{I}}^{2}$$\end{document} = 54.8%, P = 0.08) (Fig. 2F). Subgroup analyses were not conducted for NO due to the limited number of included studies considering this variable. Effect of synbiotic supplementation on GPX concentration GPx concentrations were reported in four studies (four treatment arms) including 208 participants (105 in the synbiotic group and 103 in the control group). The random-effects meta-analysis showed that synbiotic supplementation was associated with a significant increase in GPx levels (SMD: 1.40, 95% CI: 0.39 to 2.40, P < 0.01) (Fig. 2G), despite substantial heterogeneity among studies (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{I}}^{2\:}$$\end{document}= 89.7%, P < 0.001). Subgroup analyses were not performed for GPx due to insufficient data on relevant covariates across the limited number of included studies. Effect of synbiotic supplementation on OSI level OSI data were collected from two studies, which included two treatment arms and a total of 160 participants (synbiotic = 80, and control = 80). Pooled analysis using a random-effect model demonstrated a statistically significant reduction in OSI levels following synbiotic supplementation (SMD: -0.46, 95% CI: -0.78 to -0.15, P < 0.01), with no significant heterogeneity observed (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{I}}^{2}$$\end{document} = 0.0%, P = 0.58) (Fig. 2H). Subgroup analyses were not feasible for OSI because the small number of eligible studies limited statistical power. Sensitivity analysis Based on the sensitivity analyses, no single study exerted a significant influence on the overall effect estimates for TAC, MDA, or NO. In contrast, the pooled estimate for TOS was sensitive to the inclusion or exclusion of the studies by Zolghadrpour et al. (2024) and Bakhshimoghaddam et al. (2018), both of which produced substantial changes in effect size. Likewise, GSH results were affected by the studies by Zhang et al. (2025), Soleimani et al. (2019), and Kleniewska et al. (2017), whose inclusion materially altered the pooled estimate. GPx estimates varied notably when Zolghadrpour et al. (2024) and Bazyar et al. (2020) were omitted or included. SOD findings were sensitive to the studies by Bazyar et al. (2020), Ghafouri et al. (2019b), and Navaei et al. (2019). Finally, OSI demonstrated sensitivity to the inclusion of the study by Ghavami et al. (2021). Publication bias Evidence of potential publication bias was observed for MDA, TAC, and GPx. Specifically, Begg’s rank correlation and Egger’s linear regression tests both indicated bias for MDA (Egger’s test, P = 0.004; Begg’s test, P = 0.006). Regarding TAC, Egger’s test suggested the presence of small-study effects (P = 0.028), although Begg’s test was not significant (P = 0.150). For GPx, Egger’s test showed significant publication bias (P = < 0.001), although Begg’s test was not significant (P = 0.08). No significant bias was found for any of the other biomarkers, as both Begg’s and Egger’s tests were non-significant. The corresponding funnel plots are presented in Fig. S1. Discussion This systematic review and meta-analysis of 18 randomized controlled trials involving adults with diverse health conditions synthesized evidence on the effects of synbiotic supplementation on a comprehensive panel of oxidative stress biomarkers. Pooled analyses using a random-effect model demonstrated significant improvements in antioxidant defense mechanisms. These included modest increases in TAC and GSH levels, alongside more substantial elevations in SOD and GPx activities. Oxidative damage markers, including MDA, TOS, and OSI, were also reduced, while NO remained unaffected. In contrast to the earlier syntheses, the present review incorporated more recent pertinent trials. Numerous meta-analyses have investigated synbiotic effects on oxidative stress biomarkers, predominantly within specific disease populations. In contrast, the present study encompasses both healthy individuals and those with diverse pathological conditions. The present results are consistent with previous meta-analyses that documented beneficial effects on TAC, GSH, and MDA [9, 16, 17, 28–30, 80, 81]. Moreover, these findings extend the evidence to include reductions in TOS and OSI, along with a stronger elevation in GPx than observed in prior syntheses. The absence of an effect on NO contrasts with findings reported in general adult and diabetic populations [16, 17, 30]. Synbiotics may enhance gut microbiota composition and attenuate oxidative stress by modulating microbial populations and reducing lipopolysaccharide (LPS) concentrations. Recent evidence indicated that probiotic bacteria can suppress nuclear factor kappa B (NF-κB) activation, thereby inhibiting the transcription of numerous pro-inflammatory genes and mitigating oxidative stress [82]. Pronounced SOD and GPx gains could possibly reflect strain-specific induction. Specific strains, such as Lactobacillus casei, Lactobacillus fermentum [83, 84], Bifidobacterium, and Lactococcus, elevate SOD activity and GSH production [85, 86]. Certain genera produce GSH directly [85, 87–91]. In states of elevated oxidative burden, such as critical illness or psychological stress, GSH levels decline precipitously due to increased consumption and rising superoxide and hydrogen peroxide [92–97]. This depletion is exacerbated when methionine or cysteine availability is limited, as both are essential precursors for GSH synthesis via the transsulfuration pathway [92, 93]. Prolonged, high-dose synbiotic administration could mitigate this decline by enhancing GSH regeneration and reducing oxidative demand. Although a significant reduction in MDA levels was observed, the certainty of evidence was very low due to substantial heterogeneity and potential publication bias. Therefore, this finding should be interpreted cautiously. The observed reduction may be partly explained by ROS-scavenging bioactive peptides and improvements in lipid metabolism associated with synbiotic supplementation [98–100]. Although synbiotic supplementation was associated with a reduction in MDA levels, no significant effect was observed for NO. Therefore, the decrease in lipid peroxidation markers cannot be attributed to alterations in NO bioavailability [16]. Prior research indicated that NO scavenges peroxides and hydroxyl radicals, converting them into stable, inactive compounds and thereby maintaining endothelial NO bioavailability [6]. The lack of change in NO concentrations may reflect its complex physiological regulation and the multifactorial nature of nitric oxide metabolism. In addition, variability in assay methodologies, baseline redox status, and population characteristics across included studies may have contributed to the null findings [6, 16]. Overall, the observed reduction in MDA appears to be independent of measurable changes in NO levels in the present analysis. TAC increased more in trials lasting ≥ 12 weeks, in participants aged < 50 years, and in high-quality studies. This pattern indicates that the benefits accumulate progressively as a result of persistent alterations in gut microbiota composition. In contrast, GSH, MDA, and SOD exhibited significant improvements in interventions lasting less than 12 weeks, particularly in trials of good or fair methodological quality. This could be attributable to the use of synbiotics delivered via enriched foods, which may facilitate rapid production of short-chain fatty acids and thereby elicit prompt antioxidant responses. Differences in synbiotic formulation (strain count, dose, prebiotic type) and the predominance of Iranian participants (13 of 18 studies; Table 2) contributed to high statistical heterogeneity and limited generalizability, as diet and genetics strongly influence gut microbiota composition [101–103]. Although only a small proportion of included studies (4 out of 18) exhibited a high risk of bias, the certainty of evidence varied substantially across outcomes. OSI was graded as moderate, while GSH, SOD, TOS, and NO received low certainty ratings, primarily due to considerable heterogeneity and imprecision. MDA and GPx were assigned very low certainty, reflecting marked inconsistency and evidence of publication bias. Given the marked heterogeneity of the study populations, encompassing both healthy individuals and patients with diverse pathologies, the findings should be interpreted with caution. Alterations in the human gut microbiota associated with diverse pathological conditions challenge the broad applicability of microbiome-directed therapies and advocate for individualized microbiota-modulating interventions [104, 105]. None of the included studies assessed fecal microbiota composition. Consequently, direct causal relationships between synbiotic-induced microbial alterations and changes in oxidative stress biomarkers could not be established. Nevertheless, the observed effects remain mechanistically plausible, as synbiotics promote the production of short-chain fatty acids, bioactive peptides, and methyl ketones, compounds known to exert antioxidant and anti-inflammatory actions [27]. Strengths and limitations This systematic review possesses several notable strengths. Rigorous subgroup and sensitivity analyses were conducted to explore heterogeneity and confirm that pooled estimates were not unduly influenced by individual studies. Furthermore, unlike prior syntheses, the GRADE framework was applied to systematically evaluate the certainty of evidence across all outcomes. Notably, a major methodological strength that distinguishes our work from previous meta-analyses is the rigorous adherence to data integrity standards, specifically the exclusion of studies that have been formally retracted or flagged with documented concerns. The inclusion of compromised publications in prior syntheses can significantly inflate the effect sizes or introduce spurious conclusions, potentially misleading clinical decisions. This stringent filtering process, which led to the exclusion of ten trials from our analysis, ensures the pooled effect sizes reported here were based only on randomized controlled trials that met our eligibility criteria and excluded retracted or compromised studies, thereby reducing the risk of biased estimates. Consequently, our estimates for certain biomarkers may be more conservative than those reported elsewhere, but they are less likely to be inflated by compromised primary data. Moreover, the inclusion of eight oxidative stress biomarkers exceeding the scope of prior syntheses provides a more holistic assessment of synbiotic efficacy. Several limitations must also be acknowledged when interpreting the findings of the present study. A subset of the included trials exhibited limitations in design, execution, or data analysis. The majority of trials were conducted in Iran (13 of 18), introducing potential geographic and dietary bias that limits generalizability to diverse populations. Participants encompassed both healthy individuals and those with heterogeneous pathologies (e.g., T2DM, NAFLD, migraine), complicating the interpretation of uniform treatment effects. Synbiotic formulations varied widely in strain composition, probiotic dose (10⁸–10¹⁰ CFU/day), and prebiotic type and dosage, contributing to substantial statistical heterogeneity. Only a subset of studies reported outcomes for TOS, NO, OSI, and GPx, reducing precision for these markers. Specifically, for some outcomes like NO and OSI, the meta-analysis was based on a very limited number of trials and small population cohorts. This small study effect might reduce the statistical power for these specific biomarkers. Consequently, these results represent the current state of literature but should be viewed as preliminary. Crucially, no trial incorporated fecal microbiota profiling, precluding direct attribution of biomarker changes to specific microbial shifts. Long-term safety and tolerability data were also absent. Regarding all the aforementioned limitations, further studies in various populations of different geographical areas or health conditions are warranted to confirm the final results regarding symbiotic effects on oxidative stress. Hence, the current results should be interpreted with caution. Future directions Future research should prioritize multinational, randomized controlled trials employing standardized multistrain synbiotic formulations to enhance comparability and generalizability. Integration of metagenomic sequencing is essential to elucidate direct associations between microbial community shifts and alterations in oxidative stress biomarkers. Stratification by gender, baseline gut microbiota composition, and disease severity will facilitate identification of responder subgroups. Additionally, long-term studies are warranted to investigate dose-response relationships, sustained efficacy, and safety profiles. Clinical implication The findings of the present study support the adjunctive use of synbiotics in managing oxidative stress-related disorders, including T2DM, NAFLD, cardiovascular disease, and migraine, particularly in adults under 50 years of age and with intervention durations of at least 12 weeks. Given the rapid antioxidant responses observed with enriched-food delivery formats and the cumulative benefits in longer trials, synbiotics may be prioritized in clinical protocols targeting early-stage metabolic dysregulation. However, efficacy is likely to vary by individual gut microbiota composition, underscoring the need for microbiota-guided personalization before widespread therapeutic adoption. From a clinical perspective, while the overall trend of synbiotic efficacy is positive, the evidence for NO and TAS is still emerging. Therefore, clinicians should be cautious when extrapolating these specific results to practice until more robust, large-scale clinical trials confirm these findings. Conclusion Synbiotic supplementation was associated with reductions in oxidative damage markers such as MDA, TOS, and OSI and increases in antioxidant: TAC, SOD, GSH, and GPx. However, due to the limited number of studies for outcomes like NO and OSI, these specific results should be considered preliminary and interpreted with caution. These findings support the adjunctive use of synbiotics in improving oxidative stress-related disorders, particularly in younger adults (< 50 years) and with ≥ 12-week regimens, pending microbiota-guided personalization. Finally, the exact impact of synbiotics on oxidative balance is not yet fully elucidated, warranting cautious interpretation of the results. Well-controlled, multicenter RCTs with standardized formulations and integrated microbiome profiling are needed to confirm these findings and clarify clinical. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1: Appendix S1: Figure S1. A-H. Sensitivity analysis; Table S1. GRADE profile Supplementary Material 1: Appendix S1: Figure S1. A-H. Sensitivity analysis; Table S1. GRADE profile