This study was conducted in alignment with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for systematic reviews and meta-analyses (40). The protocol for this review was prospectively registered with the International Prospective Register of Systematic Reviews (PROSPERO) under the registration number CRD42025638185.

A comprehensive literature search was systematically conducted across five electronic databases: PubMed, Embase, Web of Science, the Cochrane Library, and CINAHL. The search covered publications from January 1, 2000, to December 31, 2024, as FMT techniques was not standardized in earlier literature, and did not receive broad scientific and clinical recognition until the early 2000s (41). The search strategy incorporated a combination of terms and their variants related to “Fecal Microbiota Transplantation” and “depression”. In addition to database searches, reference lists of all eligible publications were manually screened to identify potentially relevant studies not captured during the initial search. To ensure methodological rigor and consistency, two independent reviewers screened the titles, abstracts, and full texts of all identified studies. Any disagreements encountered during the selection process were resolved through consensus or by consultation with a third reviewer. The complete search algorithms are detailed in the Appendix 1.

Eligible studies included RCTs that evaluated FMT as a therapeutic intervention for depressive symptoms, regardless of dosage or route of administration. Comparators included placebo or autologous transplantation. Studies were required to report outcomes directly related to depressive symptoms and to be published in English. Exclusion criteria comprised non-human studies, non-clinical research, case reports, conference proceedings, narrative reviews, editorials, commentaries, systematic reviews, and meta-analyses.

To ensure data accuracy and methodological reliability, two reviewers independently extracted relevant information from the original publications. A standardized, pre-specified data extraction form was employed to systematically collect study characteristics, including general study details (e.g., first author and year of publication), participant demographics and sample sizes, FMT donor attributes, intervention protocols for both experimental and control groups, reported outcomes, and any adverse events. In cases where essential information was missing, efforts were made to contact the original study authors to retrieve the necessary data. Studies that failed to provide sufficient information were excluded from the final analysis.

The methodological quality of the included trials was independently appraised by two researchers using the Cochrane Risk of Bias assessment tool (42). This tool evaluates potential sources of bias across multiple domains, such as random sequence generation, blinding procedures, completeness of outcome data, selective outcome reporting, and other possible biases. Each domain was classified as having a “low,” “unclear,” or “high” risk of bias. Any discrepancies in assessment were resolved through discussion until consensus was reached.

Meta-analytical procedures were conducted using RevMan software version 5.4.1. Given the variability in measurement instruments across studies, the standardized mean difference (SMD) was calculated using Hedges’g. Statistical heterogeneity among the included studies was assessed using the I² statistic and the Chi-square test. In light of the clinical heterogeneity inherent in pooling distinct populations, random-effects models were employed as the primary analytic approach to provide more conservative and generalizable estimates. In cases of substantial heterogeneity (p ≤ 0.1 or I² ≥ 50%), additional subgroup analyses, sensitivity analyses, or narrative synthesis were conducted to explore sources of inconsistency. The overall effect size was evaluated using the Z statistic, and statistical significance was defined as p < 0.05. A summary of sensitivity analyses comparing model choices has been included in the Supplementary Materials to ensure robustness of findings.

The certainty of evidence reflects the degree of confidence in the accuracy of estimated treatment effects derived from the available research. In this study, the strength of evidence across all reported outcomes was evaluated using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) framework (43). Following the guidelines outlined in the GRADE handbook, the overall certainty of evidence (CoE) was assessed by considering five key domains for potential downgrading: risk of bias, inconsistency, indirectness, imprecision, and publication bias. The CoE for each outcome was categorized as high, moderate, low, or very low, and the rating process was carried out using the GRADEpro online tool (accessible at http://www.gradepro.org).

An initial total of 2,280 records was identified through comprehensive searches across multiple electronic databases. After removing duplicates, 1,308 unique articles remained for further evaluation. Title and abstract screening resulted in the selection of 64 studies for full-text review, based on their potential relevance to the inclusion criteria. Of these, 51 articles were excluded following detailed assessment due to various reasons such as study design, population mismatch, or insufficient outcome reporting. Ultimately, 12 RCTs met the predefined eligibility criteria and were included in the final analysis (44–55). A detailed summary of the study selection process is illustrated in Figure 1 .

The included studies were published between 2019 and 2024, representing diverse geographic regions such as China, Australia, Canada, Finland, and the United States. The sample sizes varied notably, ranging from 9 pairs to 136 pairs of participants, with a cumulative total of 347 participants in the FMT groups and 334 in the control groups. Across these studies, FMT was investigated as an intervention for various health conditions. Specifically, five trials examined disorders of gut-brain interaction (DGBIs), all of which focused on IBS (44, 52–55), two targeted neurological disorders (48, 49), another two focused on chronic illnesses accompanied by depressive symptoms (45, 53), and one study specifically evaluated patients diagnosed with major depressive disorder (47). However, all trials uniformly assessed the effect of FMT on depressive symptoms. Across studies, the mean age of participants ranged from 32.7 to 67.2 years, and the proportion of male participants varied between 13% and 87%.

FMT was delivered through various routes, including oral capsules (44, 45, 49, 52, 53, 55), colonoscopy (51, 54), gastroscopy (50), jejunal catheter (46), transendoscopic enteral tubing (TET) (48), and rectal enema (47). Controls included placebo capsules, autologous fecal microbiota transplantation, or standard medication treatment. Treatment frequencies ranged from a single administration to repeated cycles over several weeks or months. Donor screening was rigorous across all studies, though donor age and recruitment criteria varied. Most trials restricted the use of antibiotics, probiotics, and prebiotics during the intervention period; however, some studies failed to report the bowel preparation methods employed prior to FMT (47, 50, 52, 53, 55). Measurement of depressive symptoms was accomplished using various instruments: four studies applied the Hamilton Depression Rating Scale (HAM-D) (44, 48, 52, 53), two studies utilized the Hospital Anxiety and Depression Scale (HADS) (46, 55), and the remaining six studies each selected distinct scales, such as the Montgomery–Asberg Depression Rating Scale (MADRS) (45, 47, 51), the Cantonese version of the Geriatric Depression Scale (GDS-15) (49), Beck Depression Inventory (BDI) (54), and the Self-Rating Depression Scale (SDS) (50). Follow-up durations across these trials ranged widely, from a minimum of 2 weeks up to 12 months. A comprehensive overview of the included studies’ characteristics is detailed in Table 1 .

Risk of bias was assessed in accordance with the guidelines outlined in the Cochrane Handbook for Systematic Reviews of Interventions. The detailed evaluations are presented in Figures 2 and 3 . All included studies provided explicit descriptions of their randomization procedures, with two trials (54, 55) employing block randomization techniques. The trial conducted by Fang et al. (46) was an open-label study, and thus did not implement allocation concealment or blinding procedures.

In contrast, the remaining studies incorporated appropriate methods for allocation concealment, utilized placebo-controlled designs, and ensured blinding of participants, intervention providers, and outcome assessors. Two studies (46, 54) documented participant withdrawal following randomization and provided reasons for attrition; however, they did not specify how missing data were addressed. One study did not clearly define the depression outcome, and no validated measurement tool was specified (45). The rest of the trials were judged to have a low risk of bias with respect to incomplete outcome data and selective reporting. Two studies reported insufficient sample sizes (53) and author conflicts of interest (55), introducing a potential source of bias.

Publication bias was assessed through visual inspection of a funnel plot. The distribution of studies appeared to be approximately symmetrical, suggesting no substantial indication of publication bias. A detailed depiction of the funnel plot is provided in Figure 4 .

The overall certainty of evidence ranged from low to very low across the included outcomes. A comprehensive summary of the quality ratings is provided in Table 2 .

FMT has demonstrated potential therapeutic benefits in the treatment of various psychiatric and psychological disorders (56, 57). In this meta-analysis of 12 RCTs involving 681 participants, FMT significantly alleviated depressive symptoms compared with placebo or standard pharmacological treatment, with no evidence of publication bias. Although sensitivity analyses revealed a reduction in the absolute effect size following the exclusion of studies with large sample sizes or atypical designs (from SMD = -1.21 to SMD = -0.56), the overall statistical significance remained robust (p = 0.0003). This suggests that the therapeutic impact of FMT on depressive symptoms is stable and may be effective even in studies involving smaller cohorts. Of particular interest, the findings of Jiang et al. (45) reinforce the notion that FMT exerts a sustained and progressively enhanced antidepressant effect, which is consistent with the overall trends observed in the present meta-analysis. However, this study did not provide a clear definition of the depression outcome, nor did it specify the use of a validated measurement tool.

The therapeutic benefit of FMT may be explained by its ability to correct dysbiosis, restore gut microbial diversity, and modulate the MGBA, thereby influencing neuroinflammatory processes, neurotransmitter metabolism, and host immune responses (21, 35, 58). Specifically, the enrichment of anti-inflammatory taxa and restoration of SCFA-producing bacteria following FMT could attenuate systemic inflammation and improve serotonergic signaling—both of which are implicated in the pathophysiology of depression.

Notably, the heterogeneity across included trials was substantial (I² = 91%), indicating considerable variability in observed effect sizes. This high heterogeneity is likely multifactorial, arising from differences in study populations, variability in FMT delivery routes follow-up durations, and the use of diverse depression assessment tools with differing sensitivity and specificity. Furthermore, differences in FMT preparation protocols (fresh vs. frozen stool, single vs. repeated administration), donor selection criteria, and particularly donor sources (single donor vs. pooled donors, related vs. unrelated donors) may substantially influence microbiota composition and functional capacity, thereby contributing to outcome variability (59). In addition, variations in pre-treatment and baseline microbiota assessment further complicate interpretation. For example, Tian et al. (48) reported that recipients received oral ciprofloxacin and metronidazole for five days prior to FMT, whereas other studies only restricted antibiotic use during the intervention period. Zhang et al. (44) evaluated both donor and recipient microbiota, and in the matched FMT group, donor–recipient matching was based on gut microbiota structure (40.3%), diversity (23.2%), beneficial bacteria (25.2%), and harmful bacteria (11.3%). Seven additional trials (45, 46, 48, 50, 51, 54, 55) also assessed baseline microbiota, but none systematically analyzed whether recipients’ baseline microbial composition significantly influenced FMT success or engraftment. Recent evidence underscores the importance of this issue. Porcari et al. (60) demonstrated that higher recipient microbial diversity and greater compatibility between donor and recipient microbiota were key determinants of donor strain engraftment and clinical response. None of the included trials systematically evaluated these factors, which may partly explain the variability in FMT effectiveness observed across studies. While random-effects modeling accounts for some of this variability, the interpretation of pooled estimates should be made cautiously.

In the present study, both oral capsule and direct gastrointestinal FMT delivery routes produced significant improvements in depressive symptoms. Nevertheless, differences were evident between these routes, with the direct gastrointestinal group exhibiting a larger effect size in alleviating depressive symptoms (SMD: −1.29 vs −1.06). This disparity might stem from variations in microbial colonization efficiency, the pace of gut microbiota reconstitution, and patient compliance associated with each method (36, 61, 62).

Directly delivering donor microbiota to the colonic environment via endoscopy or enema likely facilitates colonization and functional activity more effectively, possibly because the colon is the primary site for microbiota colonization (63, 64). Studies have shown that after FMT administered via gastroscopy in IBS patients, the recipients’ gut microbiota profiles shift significantly toward those of the donor, a change strongly associated with symptom improvement (65, 66). Conversely, the oral capsule route may expose the introduced microbes to gastric acid and the heterogeneous gastrointestinal environment, thereby reducing the viability and functional stability of the transplanted bacteria (67). In line with these mechanistic insights, clinical data have confirmed that the endoscopic administration group achieves superior neurotransmitter modulation: after treatment, serotonin (5-HT) and γ-aminobutyric acid (GABA) levels are significantly elevated while glutamate levels are reduced, and the magnitude of these neurochemical changes far exceeds that observed with other delivery methods (46). Furthermore, the mechanical stimulation involved in endoscopic procedures might enhance signaling along the gut–brain axis and promote vagus nerve-mediated anti-inflammatory pathway activation, thereby amplifying the modulation of depression-related inflammatory factors (35, 68). On the other hand, oral capsule FMT is more convenient and safer than enema or endoscopic approaches, and its high patient compliance makes it suitable for long-term maintenance therapy (69). With repeated administration, FMT can facilitate the gradual restoration of the recipient’s gut microbiota diversity and stability, thereby extending the duration of disease remission. Therefore, the selection of an FMT delivery route should comprehensively consider factors such as microbial viability, colonization efficiency, patient compliance, procedural invasiveness, and the feasibility of repeated administration.

However, it is important to acknowledge that this methodological subgroup analysis focused solely on the delivery route, without accounting for several other potentially influential methodological factors. These include heterogeneity in donor selection protocols, variation in stool dosage and frequency, differences in bowel preparation procedures (e.g., fasting, colonoscopy prep, antibiotic preconditioning), and concomitant use of antibiotics, probiotics, or psychotropic medications. Such factors may act as confounders and contribute to variability in FMT outcomes. Therefore, while our findings suggest a potential advantage of direct gastrointestinal delivery, these results should be interpreted with caution, and future studies are warranted to isolate and systematically assess the influence of individual methodological components on clinical efficacy.

Existing research indicates that FMT exerts its therapeutic effects through mechanisms such as modulating gut microbiota composition, restoring microbiota-gut-brain axis function, ameliorating neuroinflammation, and rectifying neurotransmitter imbalances (35). However, therapeutic outcomes have been observed to vary at different follow-up time points.

Clinical observations suggest that FMT can improve psychiatric symptoms in the short term. For instance, one study focusing on patients with functional constipation accompanied by psychiatric symptoms found that within four weeks after FMT, patients experienced significant relief in both gastrointestinal and psychiatric symptoms (70). Similarly, in patients with chronic insomnia, FMT significantly improved insomnia symptoms at the 4-week mark and also had a positive effect on co-occurring anxiety and depression (30). Animal experiments further corroborate these clinical findings. In a mouse model of depression, FMT rapidly reversed depressive-like behavior, with especially pronounced effects in mice that were unresponsive to SSRIs (71). Another study demonstrated that transplanting fecal microbiota from volunteers experiencing psychological stress and subclinical depressive symptoms into mice induced depression- and anxiety-like behaviors within a short period (72), providing inverse evidence for the gut microbiota’s rapid influence on psychiatric symptoms.

Mid-term follow-up studies have revealed that FMT’s effects can be both persistent and stable. For example, a study in patients with treatment-resistant depression reported that at 12 weeks post-FMT, depression scores were significantly reduced, and this improvement was associated with stable changes in gut microbiota composition (73). Notably, improvements in psychiatric symptoms with FMT often coincide with relief of gastrointestinal symptoms. In patients with IBS, FMT not only alleviated gastrointestinal symptoms but at mid-term follow-up it also exerted a positive effect on mood (74). This dual “gut-brain” improvement effect reinforces the crucial role of the microbiota-gut-brain axis in psychiatric disorders (75).

Research on FMT’s long-term effects is relatively limited, and available data suggest that the therapeutic outcomes may vary between individuals. Some studies have reported that improvements in depressive symptoms are sustained at six months post-FMT (76), whereas others have noted a gradual diminishment of efficacy (47, 77).

In our study, we found that while FMT significantly alleviated depressive symptoms at short- and mid-term follow-ups, its effect had waned by the long-term follow-up and was no longer statistically significant. This divergence could be attributable to several factors. First, the ecological stability of the gut microbiome may be insufficient; the exogenous microbial strains introduced via FMT might not be able to colonize the host gut in the long term, causing the microbiota composition to gradually revert to its pre-transplant state (35). Second, donor-specific microbiota–induced changes in the host serum metabolome may diminish as metabolic homeostasis is restored (46, 73, 78). Finally, current clinical practice typically employs only a single FMT session, whereas studies suggest that repeated FMT can enhance efficacy by continuously modulating gastrointestinal symptoms and maintaining gut microbial diversity. This implies that the attenuation of long-term effects may be related to an insufficient frequency of intervention (79). Additionally, the propensity for depressive symptoms to relapse could offset FMT’s initial benefits. Research indicates that even if FMT achieves symptom remission, patients may later experience a return of gut dysbiosis and neuroinflammatory activation due to environmental stressors or genetic predispositions, potentially precipitating a relapse of depression (58, 80). In summary, sustaining the long-term efficacy of FMT will likely require further optimization of microbiota transplantation strategies and exploration of combined approaches to consolidate its therapeutic effects, such as adjunctive neuromodulation or lifestyle interventions.

This study analysis demonstrated that FMT significantly improved depressive symptoms in both IBS populations and neurological/psychiatric-related conditions. Notably, the effect was numerically greater in IBS populations than in neurological/psychiatric-related conditions (SMD: −1.06 vs −0.67). Given that IBS is now classified as a disorder of DGBI (81, 82), these findings are particularly relevant. Survey data indicate that depressive symptoms in patients with DGBIs are closely associated with visceral hypersensitivity, suggesting that dysregulation of the MGB axis may be a core mechanism underlying this comorbidity (83). Gut dysbiosis may enhance FMT’s antidepressant efficacy by concurrently altering SCFA/5-HTP metabolism and exacerbating neuroinflammation.

Multiple factors could underlie this observation. First, the bidirectional regulation of the gut-brain axis is considered pivotal: FMT can modulate the gut microbiota composition or function, reduce systemic inflammation, and promote neurotransmitter synthesis, thereby concurrently improving gastrointestinal dysfunction and mood symptoms (35, 84). Moreover, individuals with IBS often exhibit compromised gut barrier function and dysbiosis. These pathological conditions can exacerbate depressive symptoms by activating vagal nerve pathways and immune-inflammatory responses (85, 86). FMT’s targeted restoration of the gut microenvironment in such cases may lead to more pronounced benefits for this subgroup (87). In addition, the potential additive effects of psychological interventions warrant attention. Studies have shown that interventions such as cognitive-behavioral therapy not only directly alleviate depressive mood but also relieve gastrointestinal symptoms by reducing visceral hypersensitivity and improving autonomic nervous regulation (88, 89). This dual mechanism may yield a synergistic effect in patients with gastrointestinal disorders. Furthermore, genetic research has revealed that depression and certain gastrointestinal diseases (e.g., IBS) share genetic loci and pleiotropic genes (90). This suggests that FMT might produce broader improvements by targeting common biological pathways. It should also be noted that even individuals who do not meet clinical diagnostic criteria for depression often exhibit subclinical depressive symptoms associated with chronic low-grade inflammation (91). Thus, FMT’s anti-inflammatory properties could play a regulatory role in such individuals as well.

Our findings contrast with those of Wang et al. (2021) (36), who reported non-significant effects of FMT on depressive symptoms when analyzed as a secondary outcome in patients with IBS. Specifically, their meta-analysis yielded no significant changes at 12 weeks (MD = −0.26, 95% CI −3.09 to 2.58), 24 weeks (MD = −2.26, 95% CI −12.96 to 8.45). Several methodological and clinical differences may explain these discrepant conclusions. First, Wang et al. employed MD values, which limit comparability across different depression rating scales, whereas our study used SMD to harmonize results across diverse tools (HAM-D, MADRS, HADS, SDS, BDI, GDS-15). Second, our analysis incorporated recently published RCTs reporting more robust antidepressant effects (44, 52), which were not included in Wang et al.’s synthesis. Collectively, these refinements may explain why our meta-analysis demonstrated significant improvements in depressive symptoms for both IBS populations and neurological/psychiatric-related conditions, thereby extending and updating the evidence base regarding FMT’s antidepressant potential.

Nonetheless, the substantial heterogeneity warrants cautious interpretation. In our study, all included studies involved the gut-brain axis and could manifest with depressive symptoms, only one trial formally assessed for MDD at baseline (47). Most trials evaluated depression as a secondary outcome or enrolled participants with subclinical depressive symptoms, in some cases with baseline depression scores already below the threshold for clinical significance. Therefore, the clinical relevance of improvements in depression scores remains uncertain. In these populations, such improvements may partly reflect amelioration of the underlying condition (e.g., IBS or neurological disorders) rather than a direct antidepressant effect of FMT.

This meta-analysis systematically evaluated the therapeutic efficacy of FMT on depressive symptoms across multiple patient populations, providing robust evidence of its broad clinical benefit. A major strength lies in the inclusion of diverse participant groups, including both patients formally diagnosed with depression and those without a clinical diagnosis but exhibiting depressive symptoms, thereby enhancing the generalizability of the findings. Additionally, subgroup analyses conducted by route of administration, disease characteristics and follow-up durations allowed for a nuanced understanding of factors affecting treatment efficacy. Finally, by highlighting the bidirectional interactions between the gut microbiota and brain function, our findings support the microbiota-gut-brain axis as a promising target for depression management, and provide a foundation for future research on personalized microbiota interventions.

This meta-analysis has several important limitations that should be considered when interpreting the findings. First, there was substantial heterogeneity in clinical populations, underlying conditions, and FMT methodologies across the included trials. Patient cohorts ranged from IBS, UC, fibromyalgia, PD, severe obesity, to MDD. These conditions differ markedly in their pathophysiology and in the mechanisms by which depressive symptoms arise, which limits the comparability and generalizability of the pooled results. Second, FMT interventions varied considerably in terms of administration route, dosing regimens, donor selection, and bowel preparation strategies. These methodological differences may influence microbiota engraftment and therapeutic outcomes. Third, depressive symptoms were assessed using heterogeneous outcome measures at follow-up time points ranging from 1 week to 12 months. This variability in measurement tools and assessment timing likely contributes to the statistical heterogeneity, and may affect the comparability of effect sizes. Fourth, most included RCTs were not primarily designed to evaluate depression as a main outcome. Only one study specifically enrolled patients with clinically diagnosed MDD, and it found no significant difference between FMT and placebo. In many other trials, depression was a secondary or exploratory endpoint, or participants exhibited only subclinical depressive symptoms. Consequently, the applicability of these findings to clinical depression populations is limited. Fifth, although subgroup analyses were comprehensive, confounding factors such as concomitant medications, psychological interventions, dietary influences, and genetic predispositions were not uniformly controlled across studies, potentially influencing treatment outcomes and introducing bias. Sixth, baseline recipient microbiota composition and antibiotic pretreatment were inconsistently reported and rarely analyzed in relation to FMT outcomes, which may substantially influence engraftment success and therapeutic efficacy. Finally, the observed clinical and methodological heterogeneity, combined with the relatively small number of trials in some subgroups, means that the synthesized results should be interpreted with caution.