The effect of postbiotics supplementation on obesity and metabolic health: a systematic review and meta-analysis of randomized control trials

Abstract

Background

The prevalence of metabolic disorders such as obesity, type 2 diabetes, and dyslipidemia has increased globally. Postbiotics as non-viable microbial products or metabolites, have recently emerged as potential modulators of metabolic health due to their anti-inflammatory and insulin-sensitizing properties. In order to obtain a better viewpoint from them, this study aimed to comprehensively investigate the effects of postbiotics on metabolic health.

Methods

This systematic review and meta-analysis, adhering to PRISMA 2020 guidelines, synthesized data from 25 RCTs assessing the effects of postbiotic supplementation on metabolic parameters. Databases including PubMed, Embase, Web of Science, and Scopus were searched up to June 2025. Outcomes included glycemic indices, anthropometric measures, lipid profiles, inflammatory markers, and blood pressure.

Results

Postbiotic supplementation significantly reduced serum insulin levels (WMD: − 2.76 µU/mL), triglycerides (TG) (–8.46 mg/dL), waist circumference (WC) (–1.47 cm), and C-reactive protein (CRP) (–0.99 mg/L). However, changes in fasting blood glucose (FBG), homeostatic model assessment for insulin resistance (HOMA-IR), HbA1c, other profile lipids, blood pressure as well as weight and body mass index (BMI) were not statistically significant. Subgroup analyses revealed more pronounced benefits in younger participants, bacterial-based formulations, and interventions longer than 8 weeks. Risk of bias was low to moderate, and no major publication bias was detected.

Conclusion

Postbiotic supplementation demonstrates modest but clinically relevant benefits on insulin sensitivity, central adiposity, TG, and systemic inflammation. These effects suggest a promising adjunctive role for postbiotics in metabolic health interventions, though further trials with standardized formulations and longer durations are warranted.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12986-025-01037-5.

Introduction

In recent years, the prevalence of metabolic disorders, including obesity, type 2 diabetes mellitus (T2DM), dyslipidemia, and non-alcoholic fatty liver disease (NAFLD), has increased dramatically. This surge is largely driven by sedentary lifestyles, poor dietary habits, and complex interactions between genetic and environmental factors [1]. Although pharmacological treatments are commonly used, they often come with side effects and may have limited long-term sustainability. Consequently, interest in nutritional and microbiota-targeted interventions has grown significantly [2].

Among these interventions, postbiotics—a new generation of microbiota-based supplements—have gained increasing attention. Unlike probiotics, which rely on the viability and survival of live microorganisms in the gastrointestinal tract, postbiotics consist of inactivated microbial cells and their bioactive compounds [3]. These include short-chain fatty acids (SCFAs), cell wall fragments, peptides, enzymes, and other fermentation metabolites that can exert beneficial effects without the need for viable organisms [4].

Preclinical and clinical studies suggest that postbiotics may improve metabolic health through several mechanisms. These include modulation of the gut immune response, reduction of systemic inflammation, enhancement of insulin sensitivity, regulation of appetite-related hormones, and attenuation of hepatic steatosis and dyslipidemia [5]. Notably, SCFAs such as propionate and butyrate, key components of many postbiotic formulations, are known to activate G-protein coupled receptors (GPR41 and GPR43), leading to increased secretion of anorexigenic hormones like Glucagon-like peptide-1 (GLP-1) and Peptide YY (PYY), which contribute to reduced appetite and improved glycemic control [6].

Furthermore, postbiotic compounds derived from specific bacterial strains—such as Lactobacillus, Bifidobacterium, and Akkermansia muciniphila—have demonstrated potential in lowering serum low density lipoprotein-cholesterol (LDL-C), reducing body mass index (BMI), improving HOMA-IR, and attenuating liver fat content in individuals with metabolic disorders [7, 8]. However, despite these promising results, the current body of clinical evidence remains heterogeneous and sometimes conflicting. Differences in postbiotic composition, dosage, intervention duration, and population characteristics contribute to inconsistent findings across studies [9–11].

Several randomized controlled trials have investigated postbiotic supplementation, with some reporting significant metabolic improvements, while others have shown no clinically relevant effects. This inconsistency highlights the need for a comprehensive systematic review and meta-analysis to evaluate the overall impact of postbiotics on metabolic health [7, 8, 12–14]. Moreover, while postbiotics offer important advantages over probiotics, such as improved shelf stability, safety in immunocompromised populations, and ease of standardization, there is currently no comprehensive systematic review and meta-analysis that aggregates the available evidence and evaluates the overall efficacy of postbiotics in improving metabolic health indicators [15].

Given the increasing interest in postbiotics and their potential as a safe and effective nutritional therapy, it is essential to perform a rigorous systematic review and meta-analysis of existing human clinical trials. This would allow for a better understanding of the overall impact of postbiotic supplementation on metabolic health markers such as fasting glucose, insulin resistance, lipid profile, inflammatory cytokines, and anthropometric indices (e.g., BMI, waist circumference).

To date, most reviews have focused on probiotics and prebiotics, with little attention to postbiotics as a distinct category. To our knowledge, no systematic review and meta-analysis has comprehensively assessed the effects of postbiotics on metabolic health outcomes in humans. Therefore, the present study is novel in synthesizing the available evidence and providing an updated quantitative evaluation of postbiotic interventions. This work is expected to fill an important gap in the literature and guide future research and clinical applications.

Methods

Search strategy

This systematic review and meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines [16]. A comprehensive search was carried out from database inception through June 2025 in PubMed/MEDLINE, Web of Science, SCOPUS, and Embase, without restrictions on language or publication date. In addition to electronic databases, the reference lists of relevant reviews and included articles were manually screened to identify additional eligible trials. The search strategy employed a combination of Medical Subject Headings (MeSH) and Emtree terms as well as free-text keywords related to “postbiotics,” “bacterial lysates,” “SCFAs,” and various metabolic and cardiometabolic outcomes (e.g., “insulin,” “triglycerides,” “CRP,” “blood pressure,” etc.). Boolean operators AND/OR were used to combine terms related to intervention (e.g., “postbiotic*”, “SCFA*”, “bacterial lysate*”) and outcomes of interest (e.g., “glucose metabolism,” “lipid profile,” “CRP,” “IL-6,” “waist circumference”). Clinical trial filters were applied (e.g., “randomized controlled trial,” “RCT”). The full electronic search strategies for all databases are provided in Supplementary Appendix. The reference lists of the included publications and related review articles were manually screened to identify any additional eligible trials that might have been missed during the initial search.

Eligibility criteria

The inclusion criteria were defined using the PICOTS framework: [1] Population: Children older than 8 years old and adults aged ≥ 18 years, with or without metabolic abnormalities; [2] Intervention: Postbiotic supplementation, including SCFA-based or bacterial lysate-based compounds (Studies focusing on dietary sources of acetate such as vinegar interventions were excluded, as these do not fully meet the operational definition of postbiotics (non-viable microbial cells, cell components, or microbial metabolites) applied in this review); [3] Comparator: Placebo or non-postbiotic control groups; [4] Outcomes: At least one metabolic health indicator including glycemic markers [fasting blood glucose (FBG), hemoglobin A1c (HbA1c), insulin, HOMA-IR], anthropometric indices [weight, BMI, WC], lipid parameters [LDL-C, HDL-C, TG, total cholesterol (TC)], inflammatory markers [CRP, Interleukin-6 (IL-6)], or blood pressure [Systolic and Diastolic blood pressure (SBP, DBP)]; [5] Study design: Randomized controlled trials (parallel or crossover design); and [6] Time: Interventions with a minimum duration of 1 week and no predefined maximum duration, provided that outcome data at the longest available follow-up were reported. Studies were excluded if they were non-randomized, observational, animal-based, lacked control groups, used combined interventions (e.g., probiotics + postbiotics), or were protocols, reviews, or editorials.

Study selection and data extraction

Two independent reviewers screened all titles and abstracts, followed by full-text assessments for eligibility. Discrepancies were resolved by consensus or consultation with a third reviewer. Data extracted included: first author’s name, publication year, country, participant health status, age, BMI, sample size in intervention and control groups, type and duration of postbiotic intervention, and the reported means and standard deviations of relevant outcomes at baseline and post-intervention. When necessary, corresponding authors were contacted to retrieve missing data. Data extraction was performed using a pre-designed standardized form and cross-checked by a second reviewer. All extraction was based on the longest follow-up period reported.

Risk of bias and quality assessment

The methodological quality of included studies was assessed using the Cochrane Collaboration’s Risk of Bias tool (RoB 2.0) for randomized trials [17]. Domains evaluated included randomization process, allocation concealment, blinding of participants and outcome assessors, completeness of outcome data, selective reporting, and other biases. Two reviewers independently rated the risk of bias as “low,” “some concerns,” or “high,” and disagreements were resolved through discussion with a third reviewer. The overall risk of bias judgment was based on the Cochrane Handbook guidance: overall low risk if all domains were low, overall some concerns if at least one domain raised concerns but none were high, and overall high risk if at least one domain was judged high risk. Additionally, the overall certainty of evidence for each outcome was evaluated using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach, which considers study design limitations, consistency, precision, directness, publication bias, and effect magnitude. GRADE evidence profiles were generated for the main outcomes [18].

Data synthesis and statistical analysis

All statistical analyses were conducted using STATA version 14.0 (StataCorp LLC, TX, USA). Continuous outcomes were expressed as weighted mean differences (WMD) and 95% confidence intervals (CI). If standard deviations (SDs) were missing, they were estimated using standard formulas from reported standard errors (SE), confidence intervals, or p-values. For example, SDs were estimated by multiplying SE by the square root of the sample size (n), or calculated from baseline and final SDs using the following formula: SD_change = √[(SD_baseline^2 + SD_final^2) - (2 * R * SD_baseline * SD_final)], assuming a correlation coefficient (R) of 0.5 when not reported [19, 20]. When SDs were not directly reported, but p-values were available, we calculated SDs from the corresponding test statistics (e.g., t-value) or from SEs derived using the reported p-value and sample size, in accordance with the Cochrane Handbook guidance. A random-effects model (DerSimonian and Laird method) was applied to account for heterogeneity across studies. Between-study heterogeneity was assessed using the Cochran Q test and I² statistics. I² values were categorized as follows: low (< 25%), moderate (25–50%), substantial (50–75%), and considerable (>75%) [21]. Prespecified subgroup analyses were conducted based on participant age (< 50 vs. ≥50 years), intervention duration (≤ 8 vs. >8 weeks), and postbiotic type (SCFA-based vs. bacterial-based). Sensitivity analyses were performed using a leave-one-out approach to assess the robustness of pooled estimates. Publication bias was evaluated using visual inspection of funnel plots and Egger’s regression test [22]. When asymmetry was suspected, the trim-and-fill method was applied to estimate the number and impact of potentially missing studies. All p-values were two-tailed, and statistical significance was set at p < 0.05.

Results

Figure 1 illustrates the step-by-step process of study selection, including applied exclusion criteria. Initially, a total of 3014 records were identified through comprehensive searches across electronic databases. After eliminating duplicate entries, 2157 unique studies remained. Upon screening titles and abstracts, 2110 records were excluded for failing to satisfy the predefined eligibility criteria. Subsequently, 47 full-text articles were reviewed in detail during the secondary screening stage. Of these, 22 studies were excluded due to various reasons, such as lacking appropriate control groups or outcome data. Ultimately, 25 randomized controlled trials met all inclusion criteria and were incorporated into the final quantitative meta-analysis.

Fig. 1.

Flow chart of study selection process

Descriptive summary of eligible studies (Based on Table 1)

Table 1.

Characteristics of eligible studies

First author et al.	Years	Country	Population	Mean age year	Sex (male %)	Sample Size Study		Follow up of intervention (Weeks)	Type of RCTs	Intervention group	Control group	Baseline of BMI (kg/m2)
						Intervention	Control
1. Qaisar	2024	United Arab Emirates	Patients with chronic obstructive pulmonary disease	67.1	100	64	67	12	Parallel	Butyrate group received one 300 mg capsule	Placebo	23.6
2. Aalipanah	2025	Iran	Overweight or obese adults	42	36.4	22	22	8	Parallel	Lactobacillus rhamnosus postbiotic-fortified yogurt (120 g/day) Or Akkermansia muciniphila postbiotic-fortified yogurt (120 g/day)	Placebo	29.9
3. Chambers	2019	UK	Overweight or obese adults	60	25	12	12	6	Cross-over desig	20 g/day of inulin-propionate ester	Placebo	29.8
4. Coppola	2022		Pediatric Obesity	11	43	27	27	24	Parallel	Butyrate, 20 mg/kg body weight per day	Placebo	30.4
5. Chambers	2018	UK	Adults with non-alcoholic fatty liver disease (NAFLD)	50	50	9	9	6	Parallel	20 g/day of inulin-propionate ester	20 g/day inulin-control group	30
6. Fogacci	2024	Italy	Individuals with Liver Steatosis and Metabolic Syndrome	61	54	25	25	12	Parallel	Butyrate (500 mg/tablet)	Placebo	27.3
7. Naghibi	2024	Spain	Healthy adults	43.27	18.3	26	27	8	Parallel	Postbiotic Bifidobacterium longum CECT 7347 (50 mg/day)	Placebo	24.23
8. Lee	2025	Taiwan	Older adults	69.3	47	34	33	12	Parallel	Postbiotic heat-treated Lacticaseibacillus paracasei PS23 (PS23) (HT-PS23, 1 × 1010 cells per capsule, two capsules/day)	Placebo	24
9. Roshanravan	2018	Iran	Type 2 diabetes	45.87	33.3	15	15	6	Parallel	Butyrate group received one 600 mg capsule	Placebo	29.8
10. F. de Groot	2020	Netherlands	Participants with longstanding type 1 diabetes.	32.5	66.6	15	15	4	Parallel	4 g butyrate	Placebo	23.5
11. Companys	2022	Spain	Abdominally obese subjects	51	58	57	57	12	Parallel	Postbiotic (1010 colony forming units (CFUs) of heatinactivated B. animalis subsp. lactis CECT8145)	Placebo	32
12. Amiri	2024	Iran	Persons with obesity	392	50	25	25	8	Parallel	Butyrate group received one 600 mg capsule + hypo-caloric diet	Placebo + hypo-caloric diet	335
13. Verhaar	2024	Netherlands	Patients With Hypertension	59	43	12	11	4	Parallel	Butyrate group received one 3.9 g capsule	Placebo	24
14. Sugawara	2020	Japan	Pre-obese subjects	43.6	46.7	85	84	12	Parallel	Postbiotic fragmented Lactobacillus amylovorus CP1563 (CP1563): 200 mg	Placebo	27
15. Nakamura	2016	Japan	Overweight and mildly obese individuals	46.4	50	98	98	12	Parallel	Postbiotic fragmented Lactobacillus amylovorus CP1563 (CP1563): 200 mg	Placebo	28.4
16. Motei	2023	UK	Healthy females	48	0	20	10	16	Parallel	Postbiotic extract of Bifidobacterium breve BB091109 (500 mg)	Placebo	NR
17. Tougaard	2022	Denmark	Type 1 Diabetes	54	57	28	25	12	Parallel	3.6 g butyrate	Placebo	29
18. Rahimi	2024	Iran	Cerebrovascular accidents (CVAs) patients	60.64	85	56	55	1	Parallel	2000 mg dose of postbiotic supplement containing short-chain fatty acids and exopolysaccharides of Lactobacillus paracasei from Postbiotics Immune	Placebo	NR
19. Amiri	2024	Iran	Persons with obesity	39.32	50	25	25	8	Parallel	Butyrate group received one 600 mg capsule + hypo-caloric diet	Placebo + hypo-caloric diet	33.95
20. Goldis	2025	Romania	Children with inflammatory bowel disease (IBD)	11.16	50	44	44	12	Parallel	150 mg butyrate once a day	Placebo	NR
21. Wanitsuwan	2024	Thailand	Patients with previous colorectal cancer	62.8	45	10	8	24	Parallel	Postbiotic heat-treated containing Lacticaseibacillus paracasei SD1 and Lacticaseibacillus rhamnosus SD11 (1 × 108 cells per capsule)	Placebo	NR
22. Tougaard	2022	Denmark	Type 1 Diabetes	54	57	28	25	12	Parallel	3.6 g butyrate	Placebo	29
23. Roshanravan	2018	Iran	Type 2 diabetes	45.87	33.3	15	15	6	Parallel	Butyrate group received one 600 mg capsule	Placebo	29.8
24. Khosravi	2022	Iran	Type 2 diabetes	46.65	35.8	20	19	6	Parallel	Butyrate group received one 600 mg capsule	Placebo	29.09
25. Chambers	2014	UK	Overweight or obese adults	55.3	38.75	25	24	24	Parallel	10 g/day inulin-propionate ester	10 g/day inulin-control group	NR

BMI: body mass index

A total of 25 RCTs were included in the final analysis, encompassing a wide range of geographical locations including Iran, the United Kingdom, Japan, Spain, the Netherlands, Italy, Taiwan, Thailand, Romania, and the United Arab Emirates. The sample sizes varied considerably across studies, ranging from as few as 8 participants per group to larger trials with up to 98 participants per arm. The populations enrolled in these studies were diverse, including adults with overweight and obese, individuals with type 1 and type 2 diabetes, pediatrics with obese, cerebrovascular accident (CVA) patients, healthy adults, individuals with inflammatory bowel disease, metabolic syndrome, chronic obstructive pulmonary disease (COPD), and those with a history of colorectal cancer. The average participant age in the studies ranged from pediatric populations (~ 11 years) to older adults (> 69 years). A notable variation in sex distribution was observed, with some studies including only female participants and others with a balanced or predominantly male sample. The majority of trials followed a parallel design, while one study (Chambers 2019) adopted a cross-over approach. The intervention duration ranged from as short as 1 week to as long as 24 weeks, with most studies focusing on an 8 to 12-week timeframe. Regarding the intervention, studies employed a variety of postbiotic forms including butyrate capsules, heat-treated bacterial strains (e.g., Lacticaseibacillus paracasei, B. breve, B. animalis), and inulin-propionate ester, as well as fermented yogurt fortified with postbiotics. Dosages and delivery formats varied: for example, some trials used fixed-dose capsules (e.g., 300–600 mg butyrate), others applied weight-based dosing (e.g., 20 mg/kg/day), and others administered CFU-based bacterial components (up to 10¹⁰ CFU/day). Control groups received either placebo capsules or non-fermented products (e.g., plain inulin) matched in appearance and administration schedule. Baseline BMI of participants spanned a wide spectrum, ranging from normal weight (e.g., 23–24 kg/m² in healthy adults) to obese classifications (> 30 kg/m² in studies like Amiri and Companys et l). Two studies (Chambers 2018 and 2014) used active comparators (e.g., inulin without SCFA).

Risk of bias assessment (Based on Table 2)

Table 2.

Risk of bias assessment according to the Cochrane collaboration’s risk of bias assessment tool

Study, Year (reference)	Random sequence generation	Allocation concealment	Blinding of participants and personnel	Blinding of outcome assessment	Incomplete outcome data	Selective reporting	Overall assessment of risk of bias
1. Qaisar	Low	Low	High	Low	Unclear	Low	Unclear
2. Aalipanah	Low	Low	High	Unclear	Unclear	Low	Low
3. Chambers	Low	Unclear	High	Low	Unclear	Low	Unclear
4. Coppola	Low	Low	High	Unclea	Unclear	Low	Unclear
5. Chambers	Low	Low	Unclear	Low	Unclear	Low	Low
6. Fogacci	Low	Unclear	High	Low	Unclear	Low	Unclear
7. Naghibi	Low	Unclear	Unclear	Low	Unclear	Low	Low
8. Lee	Low	High	High	Low	High	Low	High
9. Roshanravan	Low	Low	High	Low	Unclear	Low	Unclear
10. F. de Groot	Low	Low	Unclear	Low	Unclear	Low	Low
11. Companys	Low	Low	High	Unclea	Unclear	Low	Unclear
12. Amiri	Low	Low	Unclear	Low	Unclear	Low	Low
13. Verhaar	Low	Unclear	Unclear	Low	Unclear	Low	Unclear
14. Sugawara	Low	Low	Unclear	Unclear	Unclear	Low	Unclear
15. Nakamura	Low	High	High	Low	Unclear	High	High
16. Motei	Low	Low	High	Low	Unclear	Low	Low
17. Tougaard	Low	Low	Unclear	Unclear	Unclear	Low	Low
18. Rahimi	Low	Low	Unclear	Low	Unclear	Low	Low
19. Amiri	Low	Low	Unclear	Unclear	Unclear	Low	Unclear
20. Goldis	Low	Unclear	High	Low	Unclear	Low	Unclear
21. Wanitsuwan	Low	Low	Unclear	Unclear	Unclear	Low	Unclear
22. Tougaard	Low	Low	High	Low	Unclear	Low	Unclear
23. Roshanravan	Low	Low	High	Low	Unclear	Low	Unclear
24. Khosravi	Low	Unclear	High	Low	Unclear	Low	Unclear
25. Chambers	Low	Low	High	Low	Unclear	Low	Unclear

The risk of bias across the included RCTs was generally acceptable, with most studies rated as low risk for random sequence generation and selective reporting. However, blinding of participants and personnel was frequently rated as high or unclear, particularly in interventions involving distinct sensory characteristics such as butyrate or fermented products, potentially introducing performance bias. Allocation concealment and incomplete outcome data were also commonly reported as unclear. Only a few trials, such as those by Lee and Nakamura, were judged as having high overall risk of bias, whereas the majority were categorized as low or unclear risk. These findings suggest moderate methodological quality overall, justifying cautious interpretation of results and supporting the need for sensitivity analyses and quality grading. Furthermore, the quality of the present meta-analysis was evaluated using the GRADE score system, resulting in a grade of 9 (very good quality).

Meta-analysis

Glucose metabolism outcomes

The meta-analysis results demonstrated that postbiotic supplementation significantly reduced serum insulin levels compared to placebo, with a weighted mean difference (WMD) of − 2.76 µU/mL (95% CI: − 5.00 to − 0.51; Pvalue = 0.016), and substantial heterogeneity across studies (I²=98.1%). However, no statistically significant changes were observed in other glycemic indicators. For FBG, the pooled WMD was − 2.79 mg/dL (95% CI: − 6.40 to 0.80; Pvalue = 0.130; I²=80.7%). Similarly, HbA1c and HOMA-IR showed comparable patterns, with WMDs of − 0.04% and − 0.48 respectively, though these reductions did not reach significance (Pvalue = 0.298 HbA1c and Pvalue = 0.099 HOMA-IR), and low and high heterogeneity was noted for HbA1c (I²=0.0%) and HOMA-IR (I²=98.5%) (Fig. 2).

Fig. 2.

Forest plots from the meta‐analysis of clinical trials investigating the effects Postbiotics on (a) glucose, (b) Insulin, (c) HbA1c, (d) HOMA-IR. WMD: weighted mean

Anthropometric measurements

In terms of anthropometric indices, the analysis indicated no meaningful differences between postbiotic and control groups regarding body weight (WMD: − 0.39 kg; 95% CI: − 1.22 to 0.43; Pvalue = 0.350; I²=0.0%) and BMI (WMD: − 0.42 kg/m²; 95% CI: − 0.82 to 0.02; Pvalue = 0.060; I²=57.3%). Nevertheless, a statistically significant reduction was observed in WC, with a pooled WMD of − 1.47 cm (95% CI: − 2.71 to − 0.23; Pvalue = 0.020; I²=65.8%) (Fig. 3).

Fig. 3.

Forest plots from the meta‐analysis of clinical trials investigating the effects Postbiotics on (a) weight, (b) BMI, and (c) WC. WMD: weighted mean

Lipid profile parameters

The effects of postbiotics on lipid markers were generally modest and not statistically significant for LDL-C (WMD: −3.81 mg/dL; 95% CI: − 11.22 to 3.60; Pvalue = 0.313; I²=83.1%), HDL-C (WMD: 0.70 mg/dL; 95% CI: − 0.49 to 1.89; Pvalue = 0.247; I²=78.9%), and TC (WMD: −1.74 mg/dL; 95% CI: − 5.46 to 1.98; Pvalue = 0.360; I²=44.7%). Interestingly, TG levels were significantly lowered following supplementation with postbiotics, with a WMD of − 8.46 mg/dL (95% CI: − 16.56 to − 0.36; Pvalue = 0.041; I²=61.1%) (Fig. 4).

Fig. 4.

Forest plots from the meta‐analysis of clinical trials investigating the effects Postbiotics on (a) cholesterol, (b) LDL-C, (c) HDL-C, d) TG. WMD: weighted mean

Inflammation and blood pressure indicators

Supplementation with postbiotics resulted in a notable decrease in CRP levels (WMD: − 0.99 mg/L; 95% CI: − 1.73 to −0.26; Pvalue = 0.008), despite high heterogeneity (I²=92.4%). In contrast, no significant changes were found in IL-6 (WMD: −0.97 pg/ml; 95% CI: −2.61, 0.67; Pvalue = 0.245; I² = 98.2%), SBP (WMD: 0.42 mmHg; 95% CI: −2.11, 2.96; Pvalue = 0.744; I² = 0.0%), and DBP (WMD: −1.26 mmHg; 95% CI: −4.42, 1.89; Pvalue = 0.432; I² = 39.2%) (Fig. 5).

Fig. 5.

Forest plots from the meta‐analysis of clinical trials investigating the effects Postbiotics on (a) CRP, (b) IL-6, (c) SBP, and (d)DBP. WMD: weighted mean

Subgroups analysis

Based on the subgroup analysis results from Supplementary Table 1, several statistically significant effects of postbiotic supplementation on metabolic health outcomes were observed. Specifically, WC was significantly reduced in studies with intervention durations of ≤ 8 weeks (WMD: − 1.85 cm; 95% CI: − 3.46 to − 0.24), and among participants aged < 50 years (WMD: − 2.02 cm; 95% CI: − 3.50 to − 0.54). Triglyceride (TG) levels significantly decreased with SCFAs-based postbiotics (WMD: − 14.76 mg/dL; 95% CI: − 25.88 to − 3.64). Insulin levels showed meaningful reductions in > 8-week interventions (WMD: − 3.81 µU/mL; 95% CI: − 6.21 to − 1.41) and in studies using bacterial-based postbiotics (WMD: − 4.43 µU/mL; 95% CI: − 7.74 to − 1.12). Additionally, CRP significantly decreased in participants aged ≥ 50 years (WMD: − 0.37 mg/L; 95% CI: − 0.73 to − 0.01), and IL-6 levels were reduced in bacterial-based postbiotic interventions (WMD: − 0.89 pg/mL; 95% CI: − 1.28 to − 0.50). These significant findings occurred within subgroups exhibiting variable heterogeneity levels, with some exceeding 80%, indicating potential variability across included studies. Notably, several subgroups showed low or non-significant heterogeneity (I² < 50%), including younger participants (< 50 years) for WC and TG, SCFAs-based interventions for TG, and bacterial-based postbiotics for IL-6 and insulin. These subgroups may account for the high overall heterogeneity observed in the main analyses, suggesting that participant age, intervention type, and duration are likely sources of variability across studies.

Sensitivity analysis

The leave-one-out method was applied to assess the influence of each individual study on the pooled effect size. The findings remained robust after sequential elimination of studies. Excluding studies with participants under 18 years did not materially alter the pooled effect estimates (Supplemental Figs. 1–4).

Publication bias

A visual review of the funnel plots indicated no apparent asymmetry, suggesting a low likelihood of publication bias concerning the effects of postbiotic interventions across the evaluated outcomes. This visual observation was further corroborated by Egger’s regression test, which revealed no statistically significant evidence of publication bias for any of the parameters analyzed. The P-values were as follows: weight (P = 0.53), BMI (P = 0.368), WC (P = 0.180), LDL cholesterol (P = 0.653), HDL cholesterol (P = 0.242), TC (P = 0.471), TG (P = 0.520), FBG (P = 0.250), insulin (P = 0.788), HbA1c (P = 0.404), HOMA-IR (P = 0.621), Il-6 (P = 0.652), CRP (P = 0.352), SBP (P = 0.453), and DBP (P = 0.652). Supplementary Figs. 5 through 8 present the corresponding funnel plots. Additionally, the trim-and-fill method revealed no missing studies, further supporting the robustness of the publication bias analysis.

Discussion

This comprehensive meta-analysis aimed to evaluate the effects of postbiotic supplementation on various metabolic health outcomes, including glycemic control, anthropometric measures, lipid profiles, inflammatory markers, and blood pressure.

The results revealed several significant yet nuanced findings that demand careful interpretation in light of existing literature and mechanistic insights. Notably, postbiotic supplementation resulted in a statistically significant reduction in serum insulin levels, indicating a potential improvement in insulin sensitivity. However, other glycemic parameters such as FBG, HbA1c, and HOMA-IR did not show significant changes. This discrepancy warrants deeper exploration. While serum insulin dropped significantly, the absence of parallel changes in FBG and HbA1c might be attributed to multiple factors. First, HbA1c represents an average of glucose concentrations over 2 to 3 months; thus, the relatively short duration of most included trials (often ≤ 8 weeks) may have been insufficient to manifest changes in HbA1c. Second, insulin sensitivity can improve independently of fasting glucose levels, especially when baseline glycemia is within normal limits. Many included studies enrolled metabolically healthy individuals or those with mild metabolic derangements, which reduces the magnitude of change detectable in FBG and HbA1c. Furthermore, high heterogeneity (I² >90%) among studies for glycemic outcomes suggests variability in intervention formulations, participant characteristics, and study designs. Importantly, subgroup analyses provided compelling evidence that longer interventions (>8 weeks) and bacterial-based postbiotics yielded greater insulin-lowering effects, pointing toward a duration- and strain-dependent response [23, 24]. From a mechanistic perspective, postbiotics—defined as non-viable microbial cells, cell components, or metabolites—can exert beneficial metabolic effects through several pathways. SCFAs, particularly butyrate and acetate, are known to enhance insulin sensitivity via G-protein-coupled receptors (GPR41 and GPR43), stimulation of GLP-1 secretion, and AMP-activated protein kinase (AMPK) activation in skeletal muscle and hepatic tissues [5, 6]. These mechanisms help reduce systemic inflammation and improve glucose uptake, even without major shifts in blood glucose values.

In terms of anthropometric outcomes, the meta-analysis found no significant effect of postbiotic supplementation on body weight or BMI, although there was a non-significant trend toward BMI reduction. However, a statistically significant reduction was observed in WC, which is a more specific indicator of visceral adiposity. These findings are consistent with earlier trials that observed reductions in central obesity without accompanying weight loss. For example, studies by Chambers et al. (2019) and Depommier et al. (2019) demonstrated that microbial metabolites, particularly butyrate and propionate, modulate appetite-regulating hormones such as GLP-1 and PYY and influence hepatic fat oxidation, contributing to selective reduction in abdominal fat [8, 25, 26]. Given that visceral adiposity is more strongly linked with metabolic syndrome and cardiovascular risk than general obesity, the reduction in WC is clinically relevant even in the absence of weight loss. The absence of significant effects on body weight and BMI could be attributed to the inclusion of relatively lean populations, lack of calorie restriction in study protocols, and short intervention durations. Notably, subgroup analyses showed that WC reduction was more pronounced in individuals under 50 years of age and in interventions ≤ 8 weeks, suggesting that younger individuals might respond more readily to metabolic modulation by postbiotics, possibly due to higher baseline metabolic flexibility [27].

With regard to lipid profiles, the findings were mixed. Postbiotic supplementation significantly reduced TG levels, while changes in LDL-C, HDL-C, and TC were not statistically significant. The reduction in TG is biologically plausible and aligns with findings from previous reviews and RCTs, such as the review by Zhang et al. (2021), which demonstrated that SCFA-producing postbiotics can inhibit hepatic lipogenesis and enhance lipid oxidation via AMPK and Peroxisome proliferator-activated receptor alpha (PPAR-α) pathways [28, 29]. SCFAs can also reduce the expression of lipogenic enzymes like SREBP-1c and stimulate adiponectin, leading to decreased very low density lipoprotein (VLDL) secretion and TG synthesis in the liver [11, 30]. Subgroup analyses further revealed that SCFA-based interventions had greater efficacy in TG reduction compared to bacterial lysates. The non-significant effects on LDL-C and HDL-C might stem from several causes. Firstly, cholesterol homeostasis is complex and tightly regulated by hepatic synthesis, intestinal absorption, and bile acid metabolism. Postbiotics may influence this system indirectly, but effects may take longer to materialize or require specific formulations. Secondly, inter-study variation in baseline lipid profiles and dietary intake could dilute measurable effects. Finally, HDL-C modulation is particularly challenging, as even pharmacologic agents struggle to produce meaningful changes in HDL-C levels [31].

A notable finding of this meta-analysis was the significant reduction in CRP levels, highlighting the anti-inflammatory potential of postbiotics. CRP is a sensitive systemic marker of inflammation and an independent predictor of cardiovascular risk. The reduction in CRP supports the hypothesis that postbiotics modulate gut-derived inflammation via improved intestinal barrier function and reduced endotoxemia [32]. Mechanistically, SCFAs and bacterial cell wall components downregulate NF-κB signaling and pro-inflammatory cytokine production [33]. Interestingly, although IL-6 levels did not change significantly in the overall analysis, subgroup results indicated that bacterial-based postbiotics led to a meaningful IL-6 reduction (WMD = − 0.89 pg/mL), suggesting a formulation-specific anti-inflammatory effect. The lack of consistent changes in IL-6 could be due to several reasons: variability in assay sensitivity, baseline inflammatory status, and limited statistical power. IL-6 is often more variable than CRP and may require longer intervention durations to show consistent downregulation. Prior studies, such as those by Zdybel et al. (2025), have also shown that microbial-based anti-inflammatory effects are more evident in individuals with elevated baseline inflammation or metabolic syndrome [9].

Regarding blood pressure, no significant effects were observed on systolic or diastolic blood pressure. This finding is not surprising, as postbiotics are unlikely to exert acute vascular effects unless mediated through significant weight loss, improved insulin sensitivity, or reduced systemic inflammation—changes that require prolonged intervention. Additionally, most included participants were normotensive at baseline, leaving minimal room for improvement. Some experimental studies suggest that SCFAs may modulate blood pressure through ACE inhibition or modulation of sympathetic nervous activity, but human evidence remains inconclusive [34].

Furthermore, although NAFLD is frequently discussed as a potential target, the current evidence is extremely scarce and does not permit a meta-analysis [10, 12].

One study recent reviews and meta-analyses have summarized preclinical (animal) evidence on postbiotics for obesity and metabolic health [35]. However, preclinical findings are not directly translatable to clinical practice due to important differences in species physiology, dosing (often much higher relative doses in animals), routes of administration, tightly controlled diets and environments, outcome definitions (histologic or tissue endpoints vs. clinical biomarkers), and differing risks of bias in preclinical literature. Therefore, we intentionally restricted our inclusion criteria to randomized controlled trials in human participants to provide clinically applicable estimates of efficacy and safety. Our meta-analysis synthesized 25 human RCTs and focuses on outcomes directly relevant to patients and clinicians (glycemic indices, anthropometrics, lipid profile, inflammatory markers, and blood pressure). This distinction explains why effect sizes reported in animal meta-analyses may be larger and not directly comparable with those observed in human trials.

Subgroup and sensitivity analyses provided further granularity. The benefits of postbiotics were more pronounced in interventions exceeding 8 weeks, among younger participants for WC and TG, and in SCFA-based formulations for TG and bacterial-based postbiotics for IL-6 and insulin. These findings underscore the importance of matching postbiotic type, dosage, and treatment duration with specific population subgroups for optimal efficacy. The robustness of our findings was confirmed in sensitivity analyses, including the exclusion of pediatric populations, which yielded results consistent with the primary analysis. The substantial heterogeneity observed across several outcomes in this meta-analysis likely arises from multiple sources. First, differences in baseline characteristics of participants, such as age, metabolic health status, and sex distribution, may have influenced the responsiveness to postbiotic supplementation. Second, variations in the type of intervention (SCFA-based vs. bacterial-based postbiotics), dosage, and delivery format are important contributors, as our subgroup analyses indicated formulation-dependent effects. Third, intervention duration also appeared to play a significant role, with longer interventions generally demonstrating stronger metabolic improvements. Additionally, methodological differences across trials, including sample size, study design (parallel vs. crossover), and blinding quality, may further explain part of the observed variability. These factors collectively suggest that the heterogeneity is partly inherent to the diversity of study populations and interventions, underscoring the need for more standardized and longer-term RCTs to clarify the true magnitude of postbiotic effects.

This meta-analysis has several strengths, including the inclusion of randomized controlled trials, robust subgroup and sensitivity analyses, and assessment of publication bias, which appeared minimal. However, limitations must be acknowledged. The diversity of postbiotic formulations and inconsistent reporting of compositional details limited our ability to compare across studies. Many included trials had small sample sizes and short durations, and dietary or lifestyle confounders were not consistently controlled. Furthermore, few studies assessed gut microbiota composition, which could have provided mechanistic insight into observed outcomes. Furthermore, wide variability in populations, intervention types, and durations provided an ideal context for subgroup analysis based on age, duration, and postbiotic form (SCFA vs. bacterial), as reported in the main analysis. Collectively, the included studies represent a robust and heterogeneous body of evidence that enhances the external validity of the meta-analysis findings. The lack of trials specifically addressing NAFLD outcomes (e.g., liver fat quantification) is a major limitation, and future studies should focus on this area. In some cases, SDs were imputed from p-values or other statistics, which may introduce uncertainty. This review was not pre-registered in PROSPERO due to time constraints, the lengthy approval process, and concerns regarding topic overlap and ethical considerations within the platform. This should be considered as a methodological limitation. Although vinegar/acetate interventions may provide mechanistic insights into SCFA effects, they were not included in our analysis because they fall outside the strict definition of postbiotics used in this review. Future reviews with broader inclusion criteria may consider integrating such studies.

In conclusion, postbiotic supplementation appears to confer modest but clinically relevant improvements in certain metabolic parameters, particularly insulin levels, triglycerides, CRP, and WC. These effects are likely mediated by microbial metabolites such as SCFAs and bacterial components acting through endocrine, inflammatory, and metabolic pathways. Future research should aim for longer intervention periods, standardized postbiotic formulations, and inclusion of microbiome and inflammatory biomarkers to elucidate underlying mechanisms and enhance clinical applicability.

Author contributions

ZLL, SWL, MHS: conception, design, statistical analysis, data collection, writing-original draft, supervision. MHS, ZLL: data collection and writing-original draft.All authors approved the final version of the manuscript.

Funding

No funding.

Data availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.