The Complex Interplay between the Gut Microbiome and Osteoarthritis: A Systematic Review on Potential Correlations and Therapeutic Approaches

The objective of this review is to systematically analyze the potential correlation between gut microbiota and osteoarthritis (OA) as well as to evaluate the feasibility of microbiota-targeted therapies for treating OA. Studies conducted from October 2013 to October 2023 were identified via a search on electronic databases such as PubMed, Web of Science, and Scopus, following established PRISMA statement standards. Two reviewers independently screened, assessed, and extracted relevant data, and then they graded the studies using the ROBINS I tool for non-randomized interventions studies and SYRCLE’s risk-of-bias tool for animal studies. A search through 370 studies yielded 38 studies (24 preclinical and 14 clinical) that were included. In vivo research has predominantly concentrated on modifying the gut microbiota microenvironment, using dietary supplements, probiotics, and prebiotics to modify the OA status. Lactobacilli are the most thoroughly examined with Lactobacillus acidophilus found to effectively reduce cartilage damage, inflammatory factors, and pain. Additionally, Lactobacillus M5 inhibits the development of OA by preventing high-fat diet (HFD)-induced obesity and protecting cartilage from damage. Although there are limited clinical studies, certain compositions of intestinal microbiota may be associated with onset and progression of OA, while others are linked to pain reduction in OA patients. Based on preclinical studies, there is evidence to suggest that the gut microbiota could play a significant role in the development and progression of OA. However, due to the scarcity of clinical studies, the exact mechanism linking the gut microbiota and OA remains unclear. Further research is necessary to evaluate specific gut microbiota compositions, potential pathogens, and their corresponding signaling pathways that contribute to the onset and progression of OA. This will help to validate the potential of targeting gut microbiota for treating OA patients.


Introduction
Osteoarthritis (OA) is a degenerative joint disease that progressively leads to the deterioration of articular cartilage [1].With 300 million people affected worldwide, OA is a significant cause of pain and disability in adults.It is classified as the 15th most significant pathology contributing to years lived with disability [1].
Inflammation is widely recognized as a significant contributor to both the development and progression of OA.The most common inflammatory mediators associated with this condition include interleukins (ILs), namely IL6 and IL1β, metalloproteinases (MMPs), and tumor necrosis factor α (TNFα) [2].Furthermore, the vascular endothelial growth factor (VEGF), the main stimulator of angiogenesis, and the activation of the innate immune system play a crucial role in the initiation and perpetuation of this inflammatory state.Angiogenesis is closely integrated processes in OA and may affect disease progression and pain.Furthermore, it has been reported that innate immunity plays a crucial role in the development of OA by activating synovial macrophages.This leads to the production of factors including ILs, which stimulate the production of MMPs and promote cartilage damage through aggrecans degradation.Thus, OA is the result of an imbalance between the anabolic and catabolic processes in the joint [3,4].It is a multifactorial pathology and age, gender, genetics, obesity, metabolic syndrome, and endocrine factors can play a synergic role [5][6][7][8].
The diagnosis and treatment of OA typically involve a combination of clinical assessment, medical history, diagnostic tests, and various therapeutic approaches.Once diagnosed, the management of OA primarily involves a comprehensive approach aimed at relieving symptoms [9,10].This approach includes physical therapy, drug therapy, and surgical intervention.Dieppe et al. proposed a management approach for osteoarthritis using a pyramid model [11].The base includes education, lifestyle advice, exercise therapy, and topical non-steroidal anti-inflammatory drugs (NSAIDs).Progressing up the pyramid, oral NSAIDs, cyclooxygenase inhibitors, and orthoses are considered.If relief is insufficient, intraarticular injectables like corticosteroids may be used.Emerging therapies, such as stem cells or regenerative treatments like platelet-rich plasma (PRP), are gaining popularity, though they are not mentioned in the Dieppe pyramid [12,13].However, none of these approaches has been shown to significantly modify disease progression.This is probably due to the numerous factors involved in the OA pathophysiology.Among these factors, the gut microbiota has gradually gained recognition as an important pathogenic factor in the development of OA [14][15][16][17], giving rise to the concept of the "gut-joint axis" that highlights the intricate interplay between gut microorganisms and joint health.Recently, two systematic reviews analyzed the crosstalk between OA and the intestinal microbiota, primarily focusing on gut microbiome composition, OA severity and pain, inflammatory factors, and intestinal permeability.However, none of these reviews explored the use of specific therapies for targeting the intestinal microbiota during OA.The intestinal microbiota comprises microorganisms that coexist symbiotically with the living organism in the gastrointestinal tract.It is mainly composed of bacteria but also includes viruses, fungi, archaea, and protozoa [18,19].The physiological composition of the intestinal microbiota is primarily defined by Firmicutes, Bacteroides, Proteus, Actinomycetes, Fusobacteria and Verruco microbia with Firmicutes and Bacteroides being the dominant phyla (Figure 1) [18,19].All these phyla can play critical roles in various physiological processes, including nutrient absorption, the maintenance of metabolic homeostasis, the development and maturation of the immune system, resistance to infections, protection against the development of systemic and mucosal immunity, and the production of neurotransmitters [18,19].However, it is important to highlight that genetics background affects the environment where the microbiota lives, impacting the availability of nutrients and how the immune system works [20].Host genetics significantly shape the microbiota, leading to different compositions among closely related species.Phylogenetically close species exhibit similar microbiota, and even within the same species, studies on twins show that monozygotic twins have a more similar microbiota than dizygotic twins, highlighting the crucial role genetics play in defining the microbiota [20].
Figure 1.Schematic representation of the interaction between gut microbiota and OA.On the left side: gut microbiota representation with the main phylum expressed, Firmicutes, Actinobacteria, Fusobacteria, Proteobacteria, Bacteroidetes, Verrucomicrobia; Firmicutes and Bacteroidetes were the most expressed.Right side: OA and healthy knee joint.The green arrow represents a good gut microbiota condition, and it is correlated with healthy cartilage, while the red arrow represents an imbalanced gut microbiota, and it is correlated with OA.
Gut microbiota dysbiosis, defined as an alteration in the diversity, structure, or function of the intestinal microbiota, can contribute to the development of various pathological conditions and diseases [21,22].Several factors and interventions can influence or regulate the microbiota population.Some keyways include dietary changes, probiotic supplementation, prebiotics, antibiotics, fecal microbiota transplantation, and lifestyle modifications.In the context of microbial dysbiosis and its association with inflammation, several proinflammatory cytokines may be involved, such as TNFα, IL1, IL6, IL17, and interferongamma [23].These cytokines are part of the intricate network of immune signaling and can have both protective and detrimental effects depending on the context [23,24].Dysregulation in their production, often triggered by microbial dysbiosis, is implicated in the pathogenesis of various inflammatory conditions and diseases, such as OA.In fact, the modulation of the microbiota has emerged as a potential factor in the effectiveness of certain anti-OA drugs [25].While drugs like NSAIDs and corticosteroids may impact the gut Figure 1.Schematic representation of the interaction between gut microbiota and OA.On the left side: gut microbiota representation with the main phylum expressed, Firmicutes, Actinobacteria, Fusobacteria, Proteobacteria, Bacteroidetes, Verrucomicrobia; Firmicutes and Bacteroidetes were the most expressed.Right side: OA and healthy knee joint.The green arrow represents a good gut microbiota condition, and it is correlated with healthy cartilage, while the red arrow represents an imbalanced gut microbiota, and it is correlated with OA.
Gut microbiota dysbiosis, defined as an alteration in the diversity, structure, or function of the intestinal microbiota, can contribute to the development of various pathological conditions and diseases [21,22].Several factors and interventions can influence or regulate the microbiota population.Some keyways include dietary changes, probiotic supplementation, prebiotics, antibiotics, fecal microbiota transplantation, and lifestyle modifications.In the context of microbial dysbiosis and its association with inflammation, several pro-inflammatory cytokines may be involved, such as TNFα, IL1, IL6, IL17, and interferon-gamma [23].These cytokines are part of the intricate network of immune signaling and can have both protective and detrimental effects depending on the context [23,24].Dysregulation in their production, often triggered by microbial dysbiosis, is implicated in the pathogenesis of various inflammatory conditions and diseases, such as OA.In fact, the modulation of the microbiota has emerged as a potential factor in the effectiveness of certain anti-OA drugs [25].While drugs like NSAIDs and corticosteroids may impact the gut microbiota, the direct influence of disease-modifying OA drugs (e.g., PRP) on the microbiota is not well established [25].Additionally, diet and lifestyle interventions, known to influence the microbiota, may complement traditional OA treatments [26,27].Several authors demonstrated that prebiotics, non-digestible dietary fibers, or compounds that serve as a food source for beneficial microorganisms in the gut can reverse the effect of a high-fat diet on OA by modulating the gut microbiota [26,27].However, the interplay between the gut microbiota and anti-OA interventions is a growing area of research, highlighting the potential for optimizing OA management through microbiota modulation.To intervene in the gut-joint axis as part of treatment requires a more profound mechanistic understanding of microbiome-host interactions and a detailed characterization of the complex community interactions involved.Therefore, we conducted a systematic review to provide clarity and address specific questions: What are the specific correlations between OA and the gut microbiota?What are the potentials of microbiota-targeted therapies used during OA?This systematic review evaluated the potential mechanisms of the association between gut microbiota and OA and tried to assess the potentials of microbiota-targeted therapies in OA.

Eligibility Criteria
The PICOS (Population, Intervention, Comparison, Outcomes, Study design) model was used to conduct this review.Studies that evaluated the relationship between gut microbiota and OA in patients and animals (Population), with or without a specific intervention/treatment (Intervention), with or without a comparison group (Comparison), and that described the relationships between gut microbiota and OA (Outcomes) in preclinical and clinical studies (Study design) were included.Studies from October 2013 to October 2023, were included in this review if they met the PICOS criteria.Articles that focused on OA but did not discuss gut microbiota, and vice versa, articles that analyzed the presence of other pathological conditions beyond OA, articles that did not specified intervention or specific therapies used and articles that did not assess the relationship between OA and gut microbiota were excluded.We also excluded studies in which data were not accessible or missing or those without an available full-text article.Furthermore, duplicates, reviews, letters, comment to the editor, protocols and recommendations, editorials, guidelines, and articles not written in English were excluded.

Search Strategies
A systematic literature review was conducted in October 2023 according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [28] (Table S1).The search was conducted on three databases: PubMed, Scopus, and Web of Science.The resulting combination of terms was used: "osteoarthritis" OR "osteoarthritides" AND "gastrointestinal microbiome" OR "gut microbiota".For each term, free words, and managed vocabulary specific to each bibliographic database were merged using the operator "OR".The combination of free vocabulary and/or Medical Subject Headings (MeSH) terms were reported in Table 1.

Selection Process
Following the removal of duplicate articles, using a public reference manager (Mendeley Desktop v. 1.19.8),potentially relevant articles were initially screened by two reviewers (LM, FS) based on their titles and abstracts.Any studies that did not meet the predefined inclusion criteria were excluded, and any uncertainties were addressed by involving a third reviewer (GG).Finally, the remaining studies were included in the final stage of data extraction.

Data Collection Process and Synthesis Methods
The process of data extraction and synthesis started with a systematic cataloguing of the details contained in the studies under review.To increase the validity of the process and to ensure that potentially relevant findings were not inadvertently overlooked during synthesis, two authors (LM, FS) undertook the extraction task.This involved the preparation of 2 tables (one for preclinical studies and one for clinical studies), in which various key elements were carefully recorded.The elements in Table 2 relating to preclinical studies included the following categories: species/age/sex/animals' number, OA model, aim, treatment, experimental groups and time, OA assessment, microbiome assessment, main results, relation between OA and gut dysbiosis, reference/year/country, and Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE) risk of bias [29].Instead, the data collected in Table 3, which focused on clinical studies, included the following aspects: study design, age/gender/number, OA assessment methods, study aim, treatment, groups, follow-up or experimental time, microbiota assessment, main results, ref./year/country, ROBINS-I risk of bias [30].

Risk of Bias Assessment
Two reviewers (LM and DC) analyzed the methodological quality of the included studies individually.In case of disagreement, they tried to reach consensus; if this failed, a third reviewer (FS) made the final decision.The methodological quality of the included in vivo studies was ensured according to the SYRCLE tool [29], which assesses the risk of bias of animal studies.The methodological quality of the included clinical studies was assessed using the ROBINS-I tool for the assessment of risk of bias in non-randomized studies of interventions [30].

Study Selection
The initial search found 370 studies.Of these, 123 were found using PubMed, 74 were found using Scopus, and 173 were found in Web of Science.Articles were uploaded to Mendeley Desktop version 1.17.9 to remove duplicates, and the resulting 78 articles were screened for title and abstract.Seventy-eight complete articles were screened to determine whether the publication met the inclusion criteria, and 38 were considered eligible for the review.Of the 38 articles eligible for the review, 24 were in vivo studies and 14 were clinical studies.The search strategy and study inclusion and exclusion criteria are shown in Figure 2.
methodological quality of the included clinical studies was assessed using the ROBINS-I tool for the assessment of risk of bias in non-randomized studies of interventions [30].

Study Selection
The initial search found 370 studies.Of these, 123 were found using PubMed, 74 were found using Scopus, and 173 were found in Web of Science.Articles were uploaded to Mendeley Desktop version 1.17.9 to remove duplicates, and the resulting 78 articles were screened for title and abstract.Seventy-eight complete articles were screened to determine whether the publication met the inclusion criteria, and 38 were considered eligible for the review.Of the 38 articles eligible for the review, 24 were in vivo studies and 14 were clinical studies.The search strategy and study inclusion and exclusion criteria are shown in Figure 2.   The review was registered to OSF registers (DOI 10.17605/OSF.IO/5JUQC).
In all the analyzed studies, the gut microbiota was detected by 16S rRNA amplification, which is a DNA sequence-based method that can identify different bacteria.

In Vivo Studies Results
Using different animal models of OA, studies evaluating oral supplementation during OA reported specific changes in the gut microbiota and metabolism.Specifically, in animals with chemically induced OA, the intragastric administration of quercetin led to a decrease in short-chain fatty acids and an increase in Lactobacillus and Ruminococcacea [31].Using the same OA animal model, it was demonstrated that oral bioactive compounds from chicken cartilage, specifically chondroitin sulfate and type II collagen peptides, were associated with a decrease in inflammatory cytokines.In a surgical OA model, EVs supplementation led to a reduction in senescent cells [33].Meanwhile, in a metabolic OA model, probiotic fiber supplementation reduced joint damage, leptin levels, lipid endotoxins, and alleviated dysbiosis [34].In the same animal model, oligofructose supplementation demonstrated a protective role against trauma-induced obesity-associated OA, which was primarily due to its effect on the gut microbiome [45].Similarly, in studies evaluating the effect of probiotic administration by using surgically induced OA model [35] and a chemically induced OA model [37], researchers found that Lactobacillus Acidophilus led to a reduction in inflammatory factors, cartilage damage, and pain [35,37].Beyond local effects in joint tissues, O'Sullivan et al. also observed the suppression of angiogenesis factors VEGF-A and VEGF-R1 in the distal colon combined with reduced inflammatory cytokines in peripheral tissues of Lactobacillus Acidophilus-treated OA mice.This indicates a close connection between the oral administration of Lactobacillus Acidophilus and its systemic impact throughout the body, affecting joint tissue homeostasis [35].In addition, Clostridium Butyricum administration reduced OA damage in one study [36].Lactobacillus M5 also inhibited the development of OA by preventing the development of HFD-induced obesity, protecting cartilage from damage, regulating adipokine levels, and modulating the composition of the gut microbiota in mice [38].
The studies investigating the association between gut dysbiosis induced by antibiotic administration (ampicillin, neomycin or antibiotic combined with a tryptophan-rich diet) and OA [27,39,40], using surgically induced or HFD OA models, showed that antibioticinduced gut dysbiosis reduced the level of lipopolysaccharide (LPS) and the inflammatory responses.They have also shown that antibiotic-induced gut microbiota dysbiosis reduces serum LPS levels and inflammatory responses, including the suppression of TNF-α and IL-6 levels.These changes may lead to a decrease in MMP-13 expression and an improvement in OA following joint injury [27,39,40].Thus, the depletion of the gut flora by antibiotics reduced the microbial products that suppress pro-inflammatory cytokines, thereby delaying the development of OA.Furthermore, it was demonstrated that adverse alterations in the intestinal microbiome, particularly in tryptophan metabolism, accelerated the development of OA through its interaction with aryl hydrocarbon receptor [40].Similarly, studies of traditional Chinese medicine therapies showed that maxibustion regulates the composition of the gut microbiota, leading to a reduction in OA cartilage damage.In addition, EA treatment of ST36 and GB34 prevents OA degradation by modulating lipid metabolism and gut microbiota [41][42][43].
A study by Chan et al. demonstrated that DOCA-induced hypertension led to the accumulation of p16INK4a+ senescent cells (SnCs) in the knee joint.This accumulation contributed to OA development [46].Captopril, an anti-hypertensive drug, was effective in removing p16INK4a+ SnCs and reducing OA damage [46].Furthermore, these changes were associated with a reduction in Escherichia-Shigella levels in the gut microbiome.Therefore, gut microbiota dysbiosis emerged as a metabolic link in chondrocyte senescence induced by DOCA-triggered hypertension [46].
The studies that evaluated the association between gut microbiota and OA, without any treatment and/or by evaluating the effect of HFD, found an association between OA and gut microbiota [43][44][45][47][48][49][50][51][52].A greater richness in gut microbiota composition was positively associated with a reduction in OA.Lactobacillus spp. was positively correlated with OA in one study and negatively correlated in another [35,44].Metanobrevibacter spp. was correlated with a higher OA score.Prevotella and Ruminococcus were negatively associated with OA in one study [36].

General Characteristics of Clinical Studies
Of the 38 studies included in this review, 14 were clinical studies (Table 3).Among these clinical studies, there was variability in the number of patients involved with the majority having an average of 75 patients.Only 3 studies had a different number of patients, with 2 having more than 10,000 patients [56] and 1 having only one patient [54].The age of the patients in the studies varied and ranged from a minimum of 18 years [54] to a maximum of 75 years [55].In seven studies, the anatomic site of OA was localized in the knee.In four studies, OA was localized in the knee with either the hand or the hip, while in four studies, the OA site was left unspecified, or multiple sites were delineated.The methods used to detect OA were X-ray combined with Kellgren and Lawrence score or WOMAC score.

Clinical Studies Results
Almost all the studies (13/14) aimed to establish a link between OA and gut microbiota composition, and several associations were identified [53][54][55][56][57][58][59][60][61][62][63][64][65].Gut microbiota composition was found to be altered in OA patients compared to healthy controls, involving changes both in composition and functionality as well as chronic low-grade inflammation.The Methanobacteriaceae family, Desulfoovibrionales order and Ruminiclostridiu 5 genus were negatively associated with OA, having a protective effect against the disease [56].Similarly, also Bacteroides, Agathobacter, Faecalli bacterium, and Roseburia showed a protective effect while Streptococcus and Enterococcus were associated with increased pain during OA.In addition, greater levels of Streptococcus spp.were significantly associated with higher WOMAC scores independently to possible confounders such as smoking, alcohol consumption and body mass index [57].Higher levels of Clostridiales and Firmicutes were also found in OA patients.Chen et al. also showed that OA patients can be discriminated from healthy controls using the multi-kingdom signatures, an analysis that identifies interactions between different bacteria, suggesting the potential of gut microbiota for predicting OA [62].
Two studies looked at the effect of probiotics on OA, and both showed a reduction in pain [27,54].In addition, one study tested the effect of EA on OA and found that EA increase Bacteroides and Agathobacter, contributing to the disease improvement [57].This increase may also offer protection against inflammation.In three studies [53,64,66], a direct relationship between OA and gut microbiota composition was not found.However, in two of these, one cross-sectional and one prospective, an increase in LPS was found in OA patients.This implies an indirect connection between OA and the gut microbiota given that LPS is generated by Gram-negative bacteria in the intestinal tract [65].

Risk of Bias Assessment
For the in vivo studies (n = 24), the risk of bias, reported in Figure 3, was high in almost all the studies.Approximately 72% of the studies did not report the method of sequence generation (n = 18).Almost all the studies (92%) reported that the groups were similar in terms of baseline characteristics (n = 23).The allocation was not adequately concealed in about 80% of the studies (n = 20), while it was concealed in four studies [32,33,40,45].Only one study reported that the animals were randomized during the experiment [49], while three studies used the housed blinding [40,43,66].One study selected assessors blinding [36], and two studies employed random outcome assessment [43,66].All studies included all the animals in the analyses (n = 24), reported the primary outcomes (n = 24) and were free of other biases that could lead to a high risk (n = 24).S2).
The assessment of risk of bias for the fourteen clinical trials included in this review is shown in Figure 4.For these studies, the risk of bias was mainly low with only one domain in some studies presenting a moderate risk, i.e., the selection of participants into the study in pre-intervention [53], missing data in post-intervention [53,54,57], measurement of outcomes in post-intervention [63,66], and selection of the reported result in postintervention [57,59,61,63,64].S2).
The assessment of risk of bias for the fourteen clinical trials included in this review is shown in Figure 4.For these studies, the risk of bias was mainly low with only one domain in some studies presenting a moderate risk, i.e., the selection of participants into the study in pre-intervention [53], missing data in post-intervention [53,54,57], measurement of outcomes in post-intervention [63,66], and selection of the reported result in post-intervention [57,59,61,63,64].
The assessment of risk of bias for the fourteen clinical trials included in this review is shown in Figure 4.For these studies, the risk of bias was mainly low with only one domain in some studies presenting a moderate risk, i.e., the selection of participants into the study in pre-intervention [53], missing data in post-intervention [53,54,57], measurement of outcomes in post-intervention [63,66], and selection of the reported result in postintervention [57,59,61,63,64].2) moderate risk of bias (the study provides sound evidence for a non-randomized study but cannot be considered comparable to a well-performed randomized trial).Details are reported in the Supplementary Material (Table S3).

Discussion
This review presents evidence from in vivo studies demonstrating that nutrients in the diet and prebiotics can improve the status of OA by modulating the microenvironment of the gut microbiota, including changes in composition and metabolism.This leads to a reduction in joint damage, senescent cells, leptin levels, lipids, endotoxins, inflammatory factors and pain levels.Among probiotics, Lactobacillus is certainly the most discussed in these papers [35,37,38,54,66].Preclinical data revealed that specifically the levels of Lactobacillus and other bacterial species (e.g., Bifidobacterium, Clostridium, Streptococcus, Bacteroides and Firmicutes) have the potential to be a crucial factor in the development of OA.The specific mechanisms that underlie this relationship remain under investigation, but the hypothesis is that the gut microbiome could impact systemic inflammation and immune responses, subsequently affecting joint health [37].In this review, only one study [47] reported an opposite correlation, where Lactobacillus levels were increased in OA.
Some preclinical studies, within the context of OA, also investigated antibiotics, which can alter the gut bacteria composition, in addition to dietary nutrients and prebiotics.These studies have investigated how changes in the gut microbiome induced by antibiotics might influence OA symptoms.Specifically, some research has indicated that antibioticinduced modifications in the gut microbiome could be associated with a reduction in OA symptoms [27,39,40].However, it is likely that the effects of antibiotics on OA may vary depending on individual factors, including the specific antibiotics used and the unique composition of the gut microbiome.Additional studies are necessary to gain a better understanding of the precise mechanisms by which the gut microbiome and antibiotics may impact OA.This could pave the way for more targeted approaches to managing OA through interventions in the gut microbiome.
Several preclinical studies also analyzed the relationship between HFD, OA, and the gut microbiota [42,44,45].HFD can lead to changes in the composition of the gut microbiota.Such diets can alter the balance of bacteria in the gut, potentially promoting the growth of pro-inflammatory microbes while reducing beneficial ones [44,45]; these changes in gut microbiota composition were linked to systemic inflammation and metabolic disturbances, which are associated with OA.Although animal studies have indicated a relationship between diet, gut microbiota and OA-related factors, research on humans is sparse.Only one study has investigated the translation of these findings, which did not show disparity in the fecal microbial communities between obese adults with OA and obese controls without OA [61].
In addition to the preclinical data, some clinical studies demonstrated that Lactobacillus leads to a decrease in various inflammatory factors as well as nociceptive mediators [27,54].Furthermore, a correlation was also demonstrated between elevated levels of Lactobacillus and a decrease in OA symptoms.Gut microbiota alterations, such as increased Clostridiales and Firmicutes levels, were also observed in OA patients.Furthermore, some specific bacteria were associated with lower risk of OA (Bifidobacterium), while others were linked to higher pain and disability scores in OA patients (Clostridium, Streptococcus, Bacteroides and Firmicutes).Despite these interesting findings, future larger clinical studies are mandatory to strengthen these concepts.This aspect was further confirmed by the search on Clini-calTrials.govfor ongoing clinical studies on "osteoarthritis" and "intestinal microbiome".The search yielded only four clinical studies (NCT05186714, NCT03985709, NCT03968770, NCT04172688).One of these studies is in the recruitment phase (NCT05186714), two have an unknown status (NCT03985709, NCT03968770), and the last one is completed, but no results have been posted (NCT04172688).In detail, one clinical study (NCT05186714) uses 16S rRNA amplicon sequencing of fecal DNA samples to compare changes in gut microbiota profiles and the quantity of Streptococcus species with OA pain; the second study investigates the association between changes in gut microbiota and the symptoms of knee and/or hip OA in Italian patients (NCT03985709).Finally, two studies evaluated the association between probiotics and gut microbiota and the intensity of OA pain (NCT03968770, NCT04172688).
A limitation of this systematic review is its descriptive approach.No meta-analysis of the included articles was performed due to the presence of statistically significant heterogeneity between them.This aspect was particularly evident in preclinical studies, where the risk of bias was high in almost all of them.In contrast, clinical data were generally of moderate to high quality and indicated a consistent correlation between OA and specific bacterial strains and phyla ratio in the gut microbiome.However, it should be noted that only a few of them have been reported, and none of them are large.To intervene in the gut-joint axis as part of treatment, a deeper mechanistic understanding of microbiome-host interactions and a detailed characterization of the intricate community interactions involved are necessary.Nevertheless, this review presents a thorough examination of the existing knowledge on the connection between gut microbiota and OA.It identifies specific areas that require further investigation: (1) exploring the mechanisms linking the gut microbiota and OA; (2) examining the shared pathways and synergies between probiotics, antibiotics, diet, and nutraceuticals in regulating the gut microbiota and OA; (3) investigating the relationship between age, gender, and OA-associated dysbiosis; (4) investigating the quantitative relationship between OA and the gut microbiota; and (5) investigating the relationship between dysbiosis, OA disease, and symptoms.

Conclusions
In this review, we have comprehensively examined the connection between gut microbiota and the progression of OA as well as the potential for altering gut microbiota in OA treatment, providing evidence for the existence of a gut-joint axis in the OA pathogenesis.Both preclinical and clinical studies provide evidence suggesting that probiotics have the potential to be advantageous for patients experiencing OA pain.This is achieved through the positive modulation of gut microbiota and the attenuation of low-grade inflammation via multiple pathways, as indicated by a growing body of research particularly in the laboratory settings.Furthermore, there is a pressing need for a more comprehensive exploration of confounding factors, particularly genetic background, sex, age, and socio-economic context.This exploration should encompass levels of physical activity, dietary composi-tion, and the use of concomitant prescribed medications.In addition to probiotics, also antibiotics, dietary interventions, and nutraceuticals demonstrated a role in regulating gut microbiota.However, given these potentially beneficial effects among antibiotics, dietary interventions, and nutraceuticals, it is imperative that further and larger-scale future studies be conducted to delve deeper into gut microbiome diversity also through next-generation sequencing, transcriptomics, and metabolomics.These discoveries may pave the way for OA treatment through targeted interventions in the gut microbiota.

Figure 2 .
Figure 2. The PRISMA flow diagram 2020 for the systematic review.Figure 2. The PRISMA flow diagram 2020 for the systematic review.

Figure 2 .
Figure 2. The PRISMA flow diagram 2020 for the systematic review.Figure 2. The PRISMA flow diagram 2020 for the systematic review.

23 Figure 3 .
Figure 3. SYRCLE's tool for assessing risk of bias in the in vivo studies.Green: low risk of bias; Red: high risk of bias.(1) Low risk of bias (the study is comparable to a well-performed randomized study); (2) High risk of bias (the study has some important problems).Details are reported in the Supplementary Material (TableS2).

Figure 3 .
Figure 3. SYRCLE's tool for assessing risk of bias in the in vivo studies.Green: low risk of bias; Red: high risk of bias.(1) Low risk of bias (the study is comparable to a well-performed randomized study); (2) High risk of bias (the study has some important problems).Details are reported in the Supplementary Material (TableS2).

Figure 4 .
Figure 4. ROBINS-I tool for assessing risk of bias in non-randomized clinical studies.Green: low risk of bias; Yellow: moderate risk of bias.(1) Low risk of bias (the study is comparable to a wellperformed randomized trial); (2) moderate risk of bias (the study provides sound evidence for a

Figure 4 .
Figure 4. ROBINS-I tool for assessing risk of bias in non-randomized clinical studies.Green: low risk of bias; Yellow: moderate risk of bias.(1) Low risk of bias (the study is comparable to a wellperformed randomized trial); (2) moderate risk of bias (the study provides sound evidence for a non-randomized study but cannot be considered comparable to a well-performed randomized trial).Details are reported in the Supplementary Material (TableS3).

Table 1 .
Combination of free-vocabulary and/or MeSH terms for the identification of studies in PubMed, Scopus and Web of Science.

Table 2 .
Main characteristics of preclinical studies included in the review.

Table 3 .
Main characteristics of clinical studies included in the review.