The following inclusion criteria were taken into consideration: (a) SRs/MAs that used probiotics for RA were eligible; (b) patients with a clear diagnosis of RA; (c) studies comparing probiotics to placebo were eligible; (d) disease activity score (DAS), swollen joints count (SJC), level of C-reactive protein (CRP) and tumor necrosis factor-α (TNF-α) were used as outcomes. The eligibility of each article was established by consensus between the two reviewers. The following exclusion criteria were taken into consideration: (a) studies unrelated to the subject; (b) conference proceedings and protocols; (c) animal experiments.

We comprehensively searched Embase, PubMed, Web of Science, Cochrane Library, and screened qualified SRs that had been released from database inception to February 12, 2024. We used a mix of free keywords and Mesh phrases to perform our search. The keywords mainly included RA, probiotics SA, and MA. The search strategy for PubMed is presented in Table 1 , which was adjusted to adapt to different databases.

Two reviewers independently conducted literature screening. All search results were imported into Endnote 20 to remove duplicates and inconsistent articles were removed based on the title and abstract. Finally, the full text was read out and eligible SRs were included. Two reviewers independently extracted the basic characteristics of eligible literature, including author, publication year, diagnostic criteria, sample size, intervention, comparison, and outcomes. Two reviewers crosschecked what was extracted, and consulted a third reviewer for any discrepancies.

Using AMSTAR-2 tool (23), the methodological quality of the included SRs/MAs was evaluated independently by two investigators. A third investigator was consulted on any disagreement. Out of the 16 items that the tool evaluates, 7 are considered essential domains (items 2, 4, 7, 9, 11, 13, and 15). Three possibilities remain for the evaluation: “Yes,” “Partially Yes,” and “No.”

Using PRISMA checklists (24), the reporting quality of the included SRs/MAs was evaluated independently by two investigators. A third investigator was consulted on any disagreement. Out of the 7 sections that the tool evaluates, 27 checklists are considered essential domains. Three possibilities remain for the evaluation: “Yes,” “Partially Yes,” and “No.”

Using GRADE system (25), the evidence quality of the included SRs/MAs was evaluated independently by two investigators. A third investigator was consulted on any disagreement. The publication bias, bias risk, inconsistency, indirectness and imprecision are used to evaluate the GRADE scoring system. Four categories are used to group the results: high, moderate, low, and very low.

A total of 122 studies were retrieved, and 94 studies were obtained after excluding duplicate studies. After reading the titles and abstract, 10 literatures were obtained. Of the remaining studies, 3 studies were excluded: conference abstracts (n = 2), not SR/MA (n = 1). Ultimately, this review includes 12 SR-MAs (13–19) that satisfied the inclusion requirements. The literature screening process is illustrated in Figure 1 .

A summary of the data extracted from the seven SRs/MAs is provided in Table 2 . These include the SRs/MAs published between 2017 and 2023. All reviews were published in English. The number of trails included in these SRs varied widely, ranging from 4 to 10, and the total number of participants ranged from 153 to 632. Interventions in the therapy group were probiotics, whereas placebo was used in the control group. Almost all SRs/MAs reached a positive conclusion.

AMASTAR-2 was used to assess the methodological quality, two included studies were rated as moderate quality, and the remaining studies ware rated as critically low quality. The key factors affecting the quality of the studies included items 2 (only two reviews registered a protocol), 4 (only two reviews used a comprehensive literature search strategy), and 7 (no reviews provided a list of excluded studies and justified the exclusions). The detailed results are presented in Table 3 .

PRISMA was used to assess the reporting quality, and overall, the quality of reporting remains not fully satisfactory. The key factors affecting the quality of the studies included Q5 (only two reviews registered a protocol), Q8 (only two reviews used a comprehensive literature search strategy), Q16 (three reviews provided an additional analysis), and Q23 (three reviews provided an additional analysis). The detailed results are presented in Table 4 .

The seven SRs included 21 outcomes related to the treatment of RA with probiotics. The results showed that 16 (19.05%) were rated as low quality and 5 (80.95%) as critically low quality. The risk of bias (21/21, 100%), imprecision (21/21,100%) and inconsistency (15/21,71.43%) were the main factors in obtaining the results. The details are presented in Table 5 .

In the seven included articles, all literatures reported positive results, which indicated the prospect of treatment of RA with probiotics. However, due to the low methodological quality and evidence quality of the included SRs/MAs, the results of these studies should be treated with caution. At present, due to factors such as the pathogenesis of RA and the lack of objective diagnostic criteria for RA, the long-term clinical efficacy evaluation of RA is not satisfactory, and more studies are needed in the future to conduct in-depth analysis and discussion on the above issues. In addition, the existing clinical studies also have shortcomings, such as small sample size, short follow-up time, and inconsistent efficacy evaluation. There is still insufficient evidence to confirm the superiority of probiotics treatment for RA. In the future, it is necessary to improve the quality of the original trails and conduct more prospective clinical studies with large samples and multiple centers to provide more high-quality evidence-based evidence for clinical workers to apply probiotics treatment for RA.

No SR was rated as high quality according to AMSTAR-2 tool, and no SR reported all 27 entries according to PRISMA. The risk of bias was high of the included SRs for the following reasons: no comprehensive retrieval strategy was adopted, failure to provide a list of excluded documents with reasons, evaluation and discussion of publication bias and evidence credibility, insufficient analysis of sources of heterogeneity and their impact on results, and lack of reporting and discussion of registration options and funding sources. According to the Cochrane handbook, all systematic evaluations should report the registered protocol and registration number in the methodology section (28). Unfortunately, none of the included studies declared that they had registered the protocol and provided the registration number, so we determined that they did not have a prior registered protocol, which increases the risk of bias while decreasing the transparency of the study (29). In addition, additional analyses, such as subgroup analysis, sensitivity analysis, and publication bias assessment, were not performed in the included studies when performing MA, which means that the efficacy of different probiotics for different populations of RA has not been precisely explored, and the results of current SRs are not necessarily robust and are likely to be altered by further analysis (30). According to GRADE evaluation results, all included outcomes were low or extremely low, indicating that the conclusions of the included SRs/MAs are likely to be significantly different from the actual situation. The main factors contributing to the evidence downgrade were risk of bias, imprecision and inconsistency. Almost all the original SR/MA-included trials had some defects in randomization, hiding blindness, and follow-up, and the risk of publication bias was high. Small sample sizes trails with a lack of randomization, blinding, and allocation concealment cause evidence quality to range from moderate to very low. To guarantee the availability of evidence, future SRs/MAs must be planned and carried out strictly in accordance with AMSTAR-2 and PRISMA.

Insofar as we are aware, this study offers the first thorough evaluation and synopsis of the evidence bolstering probiotic usage in RA. But there are some limitations that need to be recognized. First, the outcomes of various original studies included in the relevant SR/MA are different, and the number of original studies and sample size of some outcome indicators are small, which can lead to low evidence strength of outcome indicators. Furthermore, due to the large difference in the outcome indicators of the included studies, quantitative combination and analysis were not carried out in this study, and only the study was described according to the conclusion of the original text.