Efficacy and safety of dietary polyphenols in rheumatoid arthritis: A systematic review and meta-analysis of 47 randomized controlled trials

Objective To evaluate safety and efficacy of dietary polyphenols in the treatment of rheumatoid arthritis (RA). Methods CNKI, Pubmed, Cochrane library, Embase were searched to collect randomized controlled trials (RCTs) of dietary polyphenols in the treatment of RA. The databases were searched from the time of their establishment to November 8nd, 2022. After 2 reviewers independently screened the literature, extracted data, and assessed the risk of bias of the included studies, Meta-analysis was performed using RevMan5.4 software. Results A total of 49 records (47 RCTs) were finally included, involving 3852 participants and 15 types of dietary polyphenols (Cinnamon extract, Cranberry extract, Crocus sativus L. extract, Curcumin, Garlic extract, Ginger extract, Hesperidin, Olive oil, Pomegranate extract, Puerarin, Quercetin, Resveratrol, Sesamin, Tea polyphenols, Total glucosides of paeony). Pomegranate extract, Resveratrol, Garlic extract, Puerarin, Hesperidin, Ginger extract, Cinnamon extract, Sesamin only involve in 1 RCT. Cranberry extract, Crocus sativus L. extract, Olive oil, Quercetin, Tea polyphenols involve in 2 RCTs. Total glucosides of paeony and Curcumin involve in more than 3 RCTs. These RCTs showed that these dietary polyphenols could improve disease activity score for 28 joints (DAS28), inflammation levels or oxidative stress levels in RA. The addition of dietary polyphenols did not increase adverse events. Conclusion Dietary polyphenols may improve DAS28, reduce C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), and improve oxidative stress, etc. However, more RCTs are needed to verify or modify the efficacy and safety of dietary polyphenols. Systematic review registration https://www.crd.york.ac.uk/prospero/, identifier CRD42022315645.


Introduction
Rheumatoid arthritis (RA) is a chronic, systemic, highly disabling autoimmune disease with non-infectious, symmetrical, progressive polyarthritis as the main clinical manifestation. It primarily affects the joints, but should be considered as a syndrome that includes extraarticular manifestations, such as rheumatoid nodules, pulmonary involvement, or vasculitis, as well as systemic comorbidities (1,2). It is the most common systemic irritant arthritis and is one of the world's major public health challenges, affecting approximately 1% of the world's population (3). RA can occur at any age, and its incidence begins to increase significantly at age 25; at age 55, the incidence of rheumatoid arthritis peaks (4,5). The prevalence of RA varies among different ethnic groups. The incidence of RA is 7 percent among Native Americans and 0.2 to 0.4 percent in some other countries (6). For example, the prevalence of RA in China is 0.2% to 4% (7), the number of patients is as high as 4 million, and the remission rate is roughly only 8.6% (8). Like most other autoimmune diseases, RA is more common in women than in men in a ratio of 2-3:1 (9).
Although RA has a high disability rate and high prevalence, its pathogenesis is not fully understood. At present, the treatment of RA has entered into a comprehensive management strategy, which aims to slow down the progression of the disease, reduce the occurrence of pain and bone destruction, preserve the joint mobility of patients as much as possible, and avoid disability (10,11). The main treatments for RA include conventional synthetic antirheumatic drugs (DMARDs), glucocorticoids, non-steroidal anti-inflammatory drugs (NSAIDs), targeted synthetic DMARDs, bio-original DMARDs and biosimilar DMARDs (12, 13). Although the above drugs can significantly relieve clinical symptoms, they have a single target, many adverse reactions (such as allergic reactions and blood diseases), poor long-term efficacy, and high cost (6). Meanwhile, in the vast developing countries, due to the huge economic burden, it is currently difficult for most RA patients to receive standardized and effective treatment, which is accompanied by serious physical and mental injuries (11). As an important complementary alternative therapy, dietary supplements are now also an important adjunct option for RA patients. For example, dietary polyphenols have the characteristics of multi-target, multi-component, and multi-mechanism in pharmacology, and have definite clinical efficacy, as well as the advantages of less toxic side effects and no drug resistance (14).
Polyphenols are a diverse class of plant-derived compounds with water-soluble chemical properties (15). They are widely found in herbs, fruits, teas, red berries, coffee, red wine, and dark chocolate worldwide, and are well-known antioxidants and have been proposed as treatments for several inflammatory and metabolic disorders (16)(17)(18)(19). Studies have shown that polyphenols can prevent oxidative stress, inhibit inflammation and modulate immunity and other pharmacological effects (20), especially inhibiting inflammation and immunity. These compounds may improve fibroblast-like synoviocytes (FLS) in rats with adjuvant arthritis (AIA) by inducing inhibition of several pro-inflammatory immunochemokines and promoting apoptosis through mitochondrial signaling pathways and endoplasmic reticulum stress (21). Currently, many randomized controlled trials (RCTs) of dietary polyphenols in the treatment of RA have been published. However, the results and treatment measures of these RCTs are different, which cannot provide a basis for clinicians to formulate treatment plans for RA. Therefore, a comprehensive and in-depth summary of these RCTs is urgently needed to complement the therapeutic modalities for the treatment of RA. Hence, this study was the first comprehensive systematic review and meta-analysis of RCTs on the treatment of RA with dietary polyphenols in order to provide high-quality evidence for clinicians.
Institutions in China, ClinicalTrials.gov and Cochrane Library were searched from. The search strategy was shown in Table S1.

Inclusion and exclusion criteria 2.3.1 Participants
Participants were RA patients. The diagnosis of RA conformed to the RA diagnostic criteria in the 2010 Rheumatoid Arthritis Diagnosis and Treatment Guidelines of the Rheumatology Branch of the Chinese Medical Association or the RA diagnostic criteria proposed by the American College of Rheumatology/2017 European Federation of Rheumatology in 1987 or other recognized criteria.

Intervention
The experimental group was treated with dietary polyphenols, and the route of administration, preparation type, etc. were not limited; the treatment could be combined with conventional treatment or control group treatment. Treatment in the control group was conventional therapy or a placebo or other non-dietary polyphenol therapy (RCT researchers claimed to limit the consumption of polyphenol-rich diets in the control group.).

Study design
The study design included RCTs with no restrictions on publication time, language, quality and publication status.

Literature quality evaluation and data extraction
Two reviewers independently searched the database according to the search strategy, and initially excluded studies that did not belong to the treatment of RA with polyphenols after reading the titles and abstracts. They then independently screened literature, assessed quality, and extracted data according to inclusion and exclusion criteria, and any disagreements were resolved by brainstorming with other reviewers. The data in the literature were registered and managed in Excel form, and the extracted data included the name of the investigator, the year of publication of the literature, the general characteristics of the included patients and the number of cases, intervention measures, outcomes, etc (23). The methodological quality of the included studies was assessed according to the Cochrane Collaboration's Risk of Bias Assessment Tool (24). The main contents of the evaluation are: (1) the application of random allocation method; (2) the implementation of random concealment; (3) the implementation of blinding of research subjects and treatment plans; (4) the implementation of blinding of outcome measurers; (5) Integrity of Outcomes; (6) Selective reporting; (7) Other bias.

Statistical analysis
Meta-analysis was performed using Review Manager 5.4 provided by the Cochrane Collaboration. Enumeration data were expressed as risk ratio (RR), measurement data were expressed as weighted mean difference (WMD) or standard mean difference (SMD), and each effect size was expressed as 95% confidence interval (CI). The chisquare test was used to analyze the heterogeneity between the results. P>0.1 and I²≤50% was considered low heterogeneity; P<0.1 and I²>50% was considered high heterogeneity (25). Regardless of heterogeneity, due to the different sources of polyphenols, we use the random effects model for all outcome. Outcomes with included RCTs ≥ 4 were selected for subgroup analysis, sensitivity analysis, publication bias assessment, meta-regression analysis and GRADE score. Sensitivity analysis, publication bias assessment, metaregression were analyzed by STATA 15.0; GRADE score was assessed by the GRADEprofiler software (26).

Pomegranate extract for RA
Only Ghavipour et al., 2016 from Iran reported the results of Pomegranate extract in the treatment of RA involving 30 participants in Pomegranate extract group and 25 in control group (27). The control group received cellulose capsules as a placebo. They found that compared with placebo, pomegranate extract reduced DAS28 scores, improved joint swelling and tenderness, decreased ESR levels, and increased gluthation peroxidase (GPx) concentrations (27). This suggests that Pomegranate extract may play a role in the treatment of RA through anti-inflammatory and anti-oxidation. However, its lowering effect on MMP3, CRP and MDA has not yet been found.   Frontiers in Immunology frontiersin.org

Total glucosides of paeony for RA
A total of 23 RCTs reported total glucosides of paeony in the treatment of RA, efficacy indicators, inflammatory parameters and adverse event outcomes were reported, so a meta-analysis was performed.

Inflammatory parameters
Inflammatory parameters include CRP, ESR, RF, IL-6 and TNF-a. Frontiers in Immunology frontiersin.org

Adverse events
A total of 15 RCTs reported adverse events. Heterogeneity analysis suggested low heterogeneity among RCTs (I2 = 0%, P=0.98), and a random-effects model was used. The results showed that compared with the control group, the adverse events of the total glucosides of paeony were lower [SMD=0.55 95%CI (0.44, 0.69), P<0.00001] ( Figure 5).

Tea polyphenols for RA
Only 2 RCTs reported tea polyphenols in the treatment of RA, efficacy indicators, inflammatory parameters and adverse event outcomes were reported, so a meta-analysis was performed. Long et al. 10.3389/fimmu.2023.1024120 Frontiers in Immunology frontiersin.org

Inflammatory parameters
Inflammatory parameters include CRP and ESR.
1. Both 2 RCTs reported CRP. Heterogeneity analysis suggested high heterogeneity among RCTs (I2 = 96%, P<0.00001), and a random-effects model was used. The results showed that compared with the control group, the CRP of the tea polyphenols was lower [WMD=-1.83 95%CI (-3.08, -0.59), P=0.004] ( Figure 7). 2. Only Alghadir et al., 2016 reported ESR and they found that after the addition of tea polyphenols, the ESR was higher than control group. Adverse events in Total glucosides of paeony. DAS28 in Tea polyphenols.

Adverse events
Mirtaheri et al., 2021 showed that no adverse effects of treatment were recorded in the tea polyphenol group and the control group (57). Alghadir et al., 2016 did not report any adverse reactions, possibly because they were not monitored (58).

Puerarin for RA
Yang et al., 2018 from China reported Puerarin in the treatment of RA, involving 60 participants in experiments group and 59 participants in control group (59). They found that DAS28, ESR, CRP, IL-6 was lower after Puerarin intervention. They also found no significant difference in adverse event rates between the two groups.

Hesperidin for RA
Kometani et al., 2008 from Japan reported Hesperidin in the treatment of RA, involving 9 participants in experiments group and 10 participants in control group (60). They found that RA symptoms improved (3 out of 9 patients) after the patients drank the Hesperidin-containing beverage, compared with only 1 improvement in 10 patients in the control group.

Crocus sativus L. extract for RA
Only 2 RCTs reported Crocus sativus L. extract in the treatment of RA, efficacy indicators, inflammatory parameters, oxidative stress markers and adverse event outcomes were reported, so a metaanalysis was performed.

Inflammatory parameters
Inflammatory parameters include CRP and ESR.
1. Only Hamidi et al., 2020 reported CRP. They found that the CRP decreased after Crocus sativus L. extract intervention. CRP in Tea polyphenols.   Figure 9).

Oxidative stress markers
Only Hamidi et al., 2020 reported MDA and TAC. They found that both MDA and TAC increased after the intervention, but there was no significant difference compared with the control group (P>0.05) (62).

Adverse events
Adverse events were not reported in either RCTs, probably because they did not monitor for adverse events.

Ginger extract for RA
Only Aryaeian et al., 2019 reported Ginger extract in the treatment of RA. They found that DAS29 was reduced after the ginger intervention compared to the control group. Regarding inflammatory parameters, they found that peroxidase-activated proliferative receptor (PPAR)-g gene expression was significantly increased in the ginger group and the control group, but there was no significant difference between the two groups (63).

Cinnamon extract for RA
Only Shishehbor et al., 2018 reported Cinnamon extract in the treatment of RA. They found that DAS28 decreased, the number of tender and swollen joints decreased, and serum CRP and TNF-a levels were significantly reduced after the cinnamon extract intervention (64). However, no difference in ESR between the two groups was observed.

Sesamin for RA
Only Helli et al., 2019 reported Sesamin in the treatment of RA. They found that compared with the placebo group, after sesamin intervention, the joint pain was relieved, the number of tender joints was reduced, and the serum hs-CRP, TNF-a and cyclooxysynthase (COX)-2 levels were significantly reduced (65, 66).

Cranberry extract for RA
Only 2 RCTs reported Cranberry extract in the treatment of RA. Thimóteo et al., 2019 found that after 12 weeks of cranberry extract treatment, DAS28 was decreased, but inflammatory parameters (such as CRP, ESR, RF) were not significantly different compared with the control group (67). Fatel et al., 2021 found that ESR, CRP and DAS28 were both decreased after fish oil + cranberry extract treatment compared with the control group and the treatment group with fish oil alone (68).  ESR in Crocus sativus L.

Curcumin for RA
A total of 5 RCTs reported curcumin in the treatment of RA, efficacy indicators, inflammatory parameters and adverse event outcomes were reported, so a meta-analysis was performed.

Inflammatory parameters
Inflammatory parameters include CRP, ESR and RF.
3.19 Further analysis for outcomes 3.19

.1 Subgroup analysis
The number of RCTs for total glucosides of paeony and curcumin exceeded 4, so a subgroup analysis was performed according to the duration and dosage (Tables 2, 3).
1. Total glucosides of paeony: The difference from the results above is that although the summary results of DAS28 were statistically significant, subgroup analysis showed no statistical difference in DAS28 at each time point. The ESR results also showed lower ESR when receiving ≤12 weeks of treatment, while the effect of receiving 24 weeks of treatment was similar to that of the control group (Because there is only one RCT result at 36 weeks, it is not clear that the treatment effect of 36 weeks is better than that of the control group). In addition, for RF, patients receiving total glucosides of paeony twice a day had lower RF than controls, while patients receiving total glucosides of paeony three times a day had similar RF to controls. 2. Curcumin: The difference from the results above is that after subgroup analysis of doses, there was no statistically significant difference in ESR between the experimental and control groups regardless of doses ≤ 250 mg or 500 mg. And after subgroup analysis of time. At 8 weeks of curcumin treatment, the ESR and CRP of the experimental group were not statistically different from those of the control group, and the statistical difference only appeared at 12 weeks.

Sensitivity analysis
The sensitivity analysis was performed for total glucosides of paeony and curcumin ( Figure 13). The sensitivity analysis showed that no matter which RCTs were removed, it had little effect on the overall results, suggesting that the results were stable.

Publication bias analysis
The publication bias was performed for total glucosides of paeony and curcumin ( Figure 14). (1)  Adverse events.

Meta-regression results
The results of meta-regression showed that the duration of intervention was not the source of heterogeneity in DAS28

Evidence quality assessment
The evidence quality of total glucosides of paeony and curcumin was assessed by GRADE pro CDT (Tables 4, 5). The recommended rating for all results is low to very low.

Pomegranate extract for RA
Punica granatum L. is rich in vitamins, minerals, organic acids, proteins (79), and also rich in phenolic and flavonoid active ingredients. Pomegranate peel polyphenols mainly include gallic tannin, ellagitannin, ellagic acid, chlorogenic acid, gallic acid, catechin, epicatechin, anthocyanin, ferulic acid and quercetin (80, 81). Studies have shown that pomegranate has been consumed as a medicinal plant for thousands of years and has various properties such as anti-inflammatory, antioxidant, anti-cancer, anti-diabetic, anti-hyperlipidemic, anti-hypertensive and cardiovascular protection (82, 83). In addition, animal (84) and human (85, 86) studies did not report any serious adverse outcomes following pomegranate consumption. Animal studies have shown that ethanolic extract of pomegranate peel can significantly increase paw withdrawal latency and reduce adverse histological changes and arthritis scores (87), and reduce serum RF, MDA, IL-1b and TNFa. Karwasra et al. reported that Pomegranate extract can significantly inhibit hind paw swelling and bone destruction, and reduce complications such as erythema, improve joint inflammation, synovial hyperplasia, inflammatory cell infiltration, periarticular bone resorption, bone erosion and joint space narrowing (88).
In vitro studies indicated that pomegranate seed oil and fermented pomegranate juice extract can inhibit COX and lipoxygenase involved in triggering the inflammatory cascade (89). The main fatty acid punicic acid in pomegranate seed oil inhibits prostaglandin biosynthesis (90) and inhibits neutrophil activation and lipid peroxidation outcomes (91). In addition to the inhibitory effect of pomegranate on eicosanoid production, pomegranate also exerts its anti-inflammatory effect by inhibiting the p38-MAPK pathway and the transcription factor NF-kB (89-91). Extracts of phenolic compounds present in pomegranate peel are potential scavengers of diphenylpicrylhydrazine (DPPH) free radicals. This may be because the phenolic substances in the peel extract have strong protondonating ability and can agglomerate with hydroxyl groups to stabilize free radicals (92, 93). Morvaridzadeh et al. demonstrated that pomegranate juice caused a non-significant increase in TAC and paraoxonase and a non-significant decrease in MDA concentration in humans (94). The limitation of this systematic review is that it contains only 1 RCT, and they found that compared with placebo, pomegranate extract reduced DAS28 scores, improved joint swelling and tenderness, decreased ESR levels, and increased GPx concentrations.
Only two RCTs reported quercetin in the treatment of RA. Javadi et al., 2017 found that DAS-28 decreased and serum TNF-a levels were significantly reduced after quercetin intervention compared to placebo (28, 29). However, Bae et al., 2009 showed a negative result (30). Because meta-analyses could not be combined, the results need to be interpreted with caution.

Resveratrol for RA
Resveratrol is a natural polyphenolic antioxidant that is abundantly present in a variety of plants, especially red grape skins, and has pharmacological properties, including antioxidants, reactive oxygen species with scavenging, anticancer, and anti-inflammatory effects (120)(121)(122)(123). Previous studies have shown that resveratrol has beneficial effects on the occurrence and development of RA (124)(125)(126). Resveratrol acts directly on mitochondrial production of ROS. Its ROS scavenging potential may be driven by the dissociative junction of leaked electrons from the respiratory chain (127)(128)(129). Several studies have reported that resveratrol can reduce the activity of COX-1, COX-2, and reduce the expression of potent inflammatory parameters such as prostaglandins and leukotrienes (130,131). In addition, resveratrol reduced TNF-a and IL-1b levels in rats with adjuvant arthritis (132). Cheon et al. demonstrated that resveratrol supplementation significantly reduced inflammation, pannus formation, and cartilage damage in mice with collagen-induced arthritis, and reduced bone destruction (133).
In this systematic review, only 1 RCT reported resveratrol in the treatment of RA. Khojah et al., 2018 found that resveratrol treatment reduced swollen and tender joint counts, decreased DAS28, and decreased serum CRP, ESR, hypocarboxylated osteocalcin, MMP3, TNF-a, and IL-6. More RCTs are needed for further research in the future (31).

Garlic extract for RA
Garlic is a functional food that has been used worldwide for thousands of years. It is rich in bioactive compounds including allicin, ajoene, s-allyl cysteine, s-methyl cysteine, diallyl disulfide (DADS), diallyl sulfide (DAS), alliin, amino acids, polysaccharides and different polyphenols (134)(135)(136). The main phenolic compounds are b-resorcinol, gallic acid, pyrogallol, quercetin, rutin and protocatechuic acid (137). Garlic improves immune system function by stimulating certain cell types, such as macrophages and lymphocytes, and reduces cytokine secretion (138). The antiinflammatory effects of allicin and DAS have been demonstrated in several studies (139,140). Garlic-derived compounds can reduce serum TNF-a, IL-6 and CRP levels by inhibiting cell signaling pathways, including COX-2 and inhibit NF-kB activation (141,142). In addition, scientific studies have reported the analgesic and anti-fatigue effects of garlic (143)(144)(145). The antioxidant properties of garlic may help reduce pain in patients with rheumatoid arthritis (146). For example, garlic supplements can relieve pain in patients with knee osteoarthritis (OA) (147-150). Hussain and Salalah (149) showed that 900 mg of garlic (capsules) per day for 8 weeks significantly reduced TNF-a concentrations in patients with knee OA (17). Treatment with garlic (400 mg/kg) reduced writhing caused by PG, a dose that showed similar effects to aspirin, one of the most commonly used analgesics (33, 151,152).
Only one RCT reported Garlic extract in the treatment of RA. Moosavian et al., 2020 found that Garlic extract intervention improved antioxidant levels (increased TAC, decreased MDA), improved quality of life, and improved inflammation (decreased CRP and TNF-a), decreased joint pain and tenderness, and decreased DAS28 compared with the placebo group (32, 33).
This meta-analysis showed that total glucosides of paeony may improve the clinical manifestations of RA (decrease DAS28, ACR20 and ACR70) and inhibit inflammatory (reduce CRP, ESR, RF, IL-6 and TNF-a). Meanwhile, the addition of total glucosides of paeony may reduce the incidence of adverse events.

Tea polyphenols for RA
Tea polyphenols (such as catechins, etc.) are the most abundant in tea (166, 167). Therapeutic benefits of green tea have been seen in neurodegenerative, inflammatory, cardiovascular, and various cancers (168,169). Catechins have demonstrated their antiinflammatory effects in many studies related to pathological conditions in which inflammation is a core driver (170). Extensive in vitro studies have shown that catechins have promising applications in the treatment of RA, with differential modulation of cartilage, bone, and synovial fibroblast activity (171). In cartilage, catechins have been found to inhibit IL-1b-induced inducible NOS (iNOS) and COX-2 expression by inhibiting IkBa phosphorylation and proteasomal degradation (172,173). Catechin also inhibits IL-1b-induced phosphorylation of c-Jun, thereby preventing activating protein 1 (AP-1) from binding to DNA (174). Akhtar and Haqqi found that when IL-1b was inhibited, IL-6, IL-8 and TNF-a were also down-regulated due to the inhibition of NF-kB (175). An early study suggests that prophylactic consumption of green tea may help improve inflammation and reduce cartilage destruction associated with different forms of arthritis (176). In bone biology, catechins are thought to reduce the amount of osteoclast formation by reducing osteoblast differentiation (177). Catechin blocks the receptor activator of RANKL-mediated activation of JNK and NF-kB pathways to inhibit the expression of the transcription factor NFATc1 required for osteoclast differentiation (178). Catechins can regulate B cell activating factors belonging to the TNF family (BAFF)/PI3K/AKT/mTOR pathway to induce apoptosis, also in B lymphocytes (179). In regulating apoptosis, catechin treatment selectively downregulated Mcl-1 (anti-apoptotic protein) expression, thereby increasing the sensitivity of synovial fibroblasts to apoptosis (180). Studies have also shown that catechins can reduce the production of MMP-1, MMP-2 and MMP-3 by RA synovial fibroblasts to prevent further cartilage and bone destruction (181,182). Therefore, although tea and tea polyphenols can neutralize the inflammatory effects of IL-1b and IL-6, they also effectively utilize TNF-a to play its basic function of regulating the uncontrolled proliferation of activated synovial fibroblasts to improve the functional status of arthritis joints.
This meta-analysis showed that tea polyphenols may improve the clinical manifestations of RA (decrease DAS28, ACR20 and ACR70) and inhibit inflammatory (reduce CRP and ESR). Meanwhile, the addition of total glucosides of paeony may not increase the incidence of adverse events.

Puerarin for RA
Puerarin exists in the roots of the genus Pueraria (common name Pueraria), which is isolated from Pueraria and other species (183)(184)(185). Numerous health benefits have been attributed to puerarin, namely antioxidant (186), anti-inflammatory (187), neuroprotective (188), liver protection (189), anticancer (190), antidiabetic (191), cardioprotective (192) and anti-atherosclerotic effect (193). In terms of bone protection, puerarin is an effective compound that inhibits bone resorption and improves bone structure. It can stimulate osteoblast differentiation and inhibit osteoclastogenesis at the same time (183,194). In addition, the Ca 2+ content in the culture supernatant decreased after puerarin treatment. Puerarin can stimulate bone formation and regulate bone metabolism by inhibiting bone resorption (195). Multiple studies have found that puerarin attenuates inflammation and oxidation in mice with collagen antibody-induced arthritis through the TLR4/NF-kB signaling pathway (196). Puerarin derivative (4AC) antioxidant and inhibits TNF-a activity via MAPKs/NF-kB signaling pathway in RAW264.7 cells and collagen-induced arthritis rats (197). Puerarin inhibits inflammation and ECM degradation through the Nrf2/HO-1 axis in chondrocytes and reduces pain symptoms in osteoarthritis mice (198).

Hesperidin for RA
Hesperidin is a flavonoid that is abundant in citrus fruits. Hesperidin has many biological functions, including antioxidant, anti-inflammatory, antiviral, and anticancer activities (208). Ahmad et al. found that after treatment with hesperidin, plasma CML and IgG PTD levels were restored to 93% and 16%, respectively, through the free radical scavenging activity of hesperidin, thereby alleviating RA disease by reducing the concentration of AGEs. Therefore, the use of hesperidin may help reduce the severity of RA disease. Umar et al. found that hesperidin may inhibit collageninduced arthritis by inhibiting free radical load and reducing neutrophil activation and infiltration (209). Qi et al. (210) found that in complete Freund's adjuvant-induced arthritis in mice, hesperidin inhibited synovial cell inflammation and macrophage polarization by inhibiting the PI3K/AKT pathway. Liu et al. (211) found that hesperidin derivative 11 inhibited the proliferation of fibroblast-like synoviocytes by activating frizzled-related protein 2 secreted in adjuvant arthritis rats. Li et al. found the therapeutic effect of hesperetin on adjuvant arthritis in rats by inhibiting the JAK2-STAT3 signaling pathway (212). Hesperidin also promotes the antiinflammatory and analgesic activities of Siegesbeckia pubescens makino by inhibiting COX-2 expression and inflammatory cell infiltration (213). This systematic review showed that hesperidin may improve RA symptoms.

Crocus sativus L. extract for RA
Crocus sativus L. extract is well known in herbal medicine and has attracted the attention of researchers for its properties, especially its anti-inflammatory and proliferative abilities in bone and cartilage destructive diseases (214,215). Among these bioactive components, there are four recognized components that may be associated with the therapeutic potential of saffron, including crocin, saffron flavonoids, saffron aldehyde, and saffron (216,217). Recent studies have revealed other therapeutic and pharmacological activities of saffron (218)(219)(220)(221)(222)(223)(224)(225), such as neuroprotection, neurogenetics, antidepressant, antiapoptotic, antioxidant, and anti-inflammatory. One study found that crocin modulates serum levels of enzymatic and non-enzymatic inflammatory cytokines, including MMP-13, MMP-3, MMP-9, HAases, TNF-a, IL-1b, NF-kB, IL -6, COX-2 and PGE2 and ROS media (226). Crocin also increased levels of GSH, SOD, CAT and GST. In addition, inhibition of the exoglycosidase cathepsin-D and tartrate-resistant acid phosphatase in the bone near the joint by crocin protects bone resorption (226). Rasol et al. found that TNF-a and IL-1b levels were decreased and SOD and GR activities were increased after crocin intervention (226). Hu et al. found that paw swelling and ankle diameter were significantly reduced in crocintreated rats. Histological analysis also showed reduced inflammation in joints and other organs, such as the spleen. In addition, TNF-a and TGF-b1 levels were decreased in synovial tissue (227). In a similar study, Liu et al. found that MMP-1, MMP-3, and MMP-13 protein expression levels were decreased in RA rats after crocin intervention (228). Li et al. showed similar results, suggesting that crocin had a positive effect on RA-induced rats (229). In an in vitro study, Li et al. demonstrated that crocin at 500 μM (5,000 mg/ ml) reduced levels of TNF-a, IL-1b, and IL-6 in human FLS. Furthermore, crocin caused lower levels of p-IkBa, p-IkB kinase a/b and p65 expression, demonstrating its effect on the NF-kB pathway (230). Wang et al. showed that crocin inhibits Wnt/bcatenin and Wnt signaling pathways to reduce pain-related cytokines, and glial activation may alleviate RA-induced neuropathic pain in rats (231).
However, this meta-analysis showed that the efficacy of Crocus sativus L. extract may not be significantly different from the control group.  1 Downgraded one level due to serious risk of bias (random sequence generation, allocation concealment, blinding, incomplete outcomes) and most of the data comes from the RCTs with moderate risk of bias. 2 Downgraded one level due to the probably substantial heterogeneity. 3 Downgraded one level due to the total sample size fails to meet the optimal information size. 4 Downgraded one level due to the publication bias.

Ginger extract for RA
Ginger has been cultivated in China and other countries around the world since ancient times as a source of medicinal plants for spice and therapeutic benefits (232). The main components of ginger are 6shogaol, ginger oil terpene, water fennel, camphor terpene, gingerol, eucalyptus, etc. (232)(233)(234). Evidence suggests that consuming ginger may help relieve joint pain associated with RA (235). Kiuchi et al. discovered the potential of ginger to inhibit the synthesis of prostaglandins, which are key to inflammation. Further research found that ginger exhibits anti-inflammatory activity by inhibiting the biosynthesis of prostaglandins and leukotrienes (236). Ribel-Madsen et al. (237) conducted the in vitro anti-inflammatory effect of ginger, and ginger-treated synoviocytes showed similar inhibitory effects as betamethasone by inhibiting the production of cytokines IL-1 and IL-6. Yang et al. observed the analgesic and anti-inflammatory effects of 6-gingerol (238). Ojewole observed potential analgesic and anti-inflammatory activity of ginger, which can be used to reduce pain and inflammation caused by arthritis (239). Srivastava et al. observed the antiarthritic activity of ginger in patients who independently experienced RA, OA and muscle discomfort (240). The beneficial effect of ginger on reducing RA-related pain may be due to the inhibition of prostaglandin and leukotriene biosynthesis (240). van Breemen et al. found that 10-gingerol, 8-shogaol, and 10-shogaol strongly inhibited COX 2, thereby significantly reducing inflammation (241). In addition, ginger can inhibit the biosynthesis of leukotrienes by inhibiting 5-lipoxygenase (242). Nurtjahja-Tjendraputra et al. also demonstrated the inhibitory effect of ginger on COX-1 activity (243). One study found that components of ginger significantly inhibited the release of pro-inflammatory cytokines (IL-12, TNF-a and IL-1 b) and pro-inflammatory chemokines in LPSinduced macrophages (244). 6-Gingerol significantly inhibits Ikba phosphorylation, NF-kb nuclear activation and PKC-a translocation, which in turn inhibits Ca mobilization and disruption of mitochondrial membrane potential in LPS-stimulated macrophages, thereby inhibiting inducible nitric oxide synthase and TNF-a express and reduce inflammation (245). This systematic review showed that Ginger extract may decrease DAS28.

Cinnamon extract for RA
Cinnamon, one of the most commonly used spices in the world and one of the oldest herbal medicines used to treat certain diseases and inflammations, has been used to treat RA in China for nearly 2000 years (246). Cinnamon can modulate immune system function by modulating anti-inflammatory and pro-inflammatory gene expression (247)(248)(249). Cinnamaldehyde is the main active component of cinnamon, and its anti-inflammatory effect has been observed in several studies (250)(251)(252). Several other flavonoids, including anti-inflammatory hesperidin and quercetin, have also been extracted from cinnamon (253,254). Scientific evidence suggests that cinnamon extract can be used to modulate the immune system, as well as prevent and treat inflammation (255). The beneficial effects of cinnamon extract and its polyphenols on reducing serum levels of TNF-a, CRP and IL-6, as well as improving clinical symptoms and antioxidant activity have been reported in animal models (256). In vitro experiments also demonstrated the beneficial effects of cinnamon polyphenols in improving immune responses by modulating the expression of pro-and antiinflammatory cytokine genes (257,258). In experimental arthritis studies, cinnamaldehyde in cinnamon significantly inhibits joint disease in experimentally arthritic animals. Cinnamaldehyde can not only significantly reduce the IL-6 content of inflammatory mediator TNF-a in peripheral mononuclear cells of RA patients (259), but also inhibit the release of IL-1b and matrix MMP-13 from synovial fibroblasts in arthritis patients (260). Cinnamaldehyde can significantly reduce the levels of TNF-a, IL-6 and IL-1b in the peripheral blood of collagen-induced arthritic rats, and significantly increase the content of the anti-inflammatory factor IL-10, exerting a systemic anti-inflammatory effect (261). Liu et al. found that cinnamaldehyde may affect the production of IL-1b by inhibiting HIF-1a, and may affect the maturation of IL-1b by regulating the formation of NLRP3 inflammasome through the succinate/HIF-1a axis (262).
This systematic review showed that Ginger extract may decrease DAS28, the number of tender and swollen joints, and serum CRP and TNF-a levels.

Sesamin for RA
Sesame is an important traditional health food that has been used in Asian countries for thousands of years to improve nutritional status and prevent various diseases (263). Sesame seeds contain high amounts of oil and protein, as well as various lignans (such as sesamin) (264,265). Several studies have shown that sesamin (one of the most abundant lignans in sesame) has various physiological functions, including antioxidant, antihypertensive, anti-obesity and hypolipidemic effects (266)(267)(268)(269). The anti-inflammatory properties of sesame compounds have been reported in rat models (270). The results of another study showed that sesamin inhibited lipopolysaccharide-induced inflammation by inhibiting p38 mitogen-activated protein kinase and NF-kB, which are the major pathways regulating cytokine production. Based on this result, sesamin may also prevent cartilage degeneration in other joint diseases such as RA (271,272).
This systematic review showed that sesamin may improve the joint pain, decrease the number of tender joints, and decrease serum hs-CRP, TNF-a and COX-2 levels.

Cranberry extract for RA
Cranberry (Vaccinium macrocarpon) juice has strong antioxidant activity, mainly containing polyphenolic compounds such as flavonols (myricetin and quercetin), anthocyanins and procyanidins (273). Clinical studies have shown that cranberry juice has beneficial effects on biomarkers of cardiovascular disease risk (274,275). Several intervention studies have found that cranberry has beneficial effects on biomarkers of oxidative stress, dyslipidemia, and inflammation in healthy people (276,277) and in patients with type 2 diabetes (278) and metabolic syndrome (279). Studies have found that quercetin (flavonol), which is abundant in cranberry, can lead to a significant down-regulation of the nuclear factor kappa B (NF-kB) pathway (280). Additionally, resveratrol has been shown to be another abundant polyphenol in cranberries. It can inhibit the expression of inflammatory genes related to cardiovascular disease by activating NF-kB and Janus kinase/signal transducer and transcriptional pathway activator in cultured cells (273). NF-kB regulates the expression of many pro-inflammatory genes, including adhesion molecules, IL-6, and TNF-a. Other components present in cranberries, such as proanthocyanidins, anthocyanins, hydroxycinnamic acid, and acetylsalicylic acid, inhibit NF-kB activation (275).
This systematic review showed that Cranberry extract may decrease DAS28, and serum ESR and CRP.

Olive extracts and oil for RA
The current study found that olive oil contains high amounts of polyphenolic compounds. Numerous studies have found polyphenolic extracts of olive oil as antioxidants to prevent and treat cardiovascular disease and prevent certain types of cancer, as well as reduce the incidence of coronary heart disease and stroke (281)(282)(283). Furthermore, it has been shown to inhibit IL-1b-induced MMP, TNF-a and IL-6 production in the SW982 human synovial fibroblast cell line (284). Polyphenol extracts also down-regulated COX-2 and mPGE-1 induced by IL-1b. In addition, it inhibits MAPK phosphorylation and NF-kB activation (285, 286). This systematic review showed that Olive oil may decrease DAS28, relieve joint pain and decrease the number of painful joints and the number of swollen joints. The study also found that a Mediterranean diet centered on olive oil can reduce the risk of autoimmune diseases such as RA (287, 288), which is strong evidence in the treatment of RA.

Curcumin for RA
Curcumin, a yellow pigment and active component of turmeric (Curcuma longa), is one of the well-known natural compounds with a wide range of pharmacological activities and potential immunomodulatory properties (289,290). Curcumin effectively inhibits the production of several pro-inflammatory mediators by mediating multiple inflammatory signaling pathways (such as JAK-STAT signaling pathway/P38 MAPK signaling pathway, NF-kB signaling pathway, etc.), including TNF-a, IL-1b, IL-6, IL-12 and IL-8 (291-294). Curcumin has a high ability to modulate inflammation in RA by suppressing pro-inflammatory immune cell populations and suppressing the production of immuneinflammatory cytokines and chemokines (295,296). Studies have found that curcumin can significantly block the expression of IL-6 in FLS stimulated by IL-1b in RA patients (297). Curcumin can also reduce the protein expression levels of IL-6, IL-8, MCP-1, MMP-1 and MMP-3 in FLS of RA patients (298). Several studies on animal models of RA also confirmed this finding. Curcumin can reduce the production of pro-inflammatory cytokines, including TNF-a, IL-1b, IL-6 and MCP-1 (299-304). Curcumin has also been found to inhibit the activation, proliferation and differentiation of naive CD4+ T cells into T helper (TH) 1 and TH 17 subtypes. These two cells play a key role in the pathogenesis of RA by producing key pro-inflammatory cytokines including TNF-a, IFN-g, IL-17, IL-21 and IL-23 and are responsible for joint and bone destruction (305,306). In addition, curcumin has a strong ability to induce regulatory T cell (Treg) differentiation and inhibit TH1-and TH17-mediated inflammatory responses (306). This meta-analysis showed that curcumin may decrease DAS28, CRP, ESR and RF.

Mechanisms of other dietary polyphenols with medicinal potential
Dietary polyphenols are currently widely used as alternative therapies, and preclinical and clinical studies have demonstrated their efficacy in the treatment of RA. In addition to the 15 dietary polyphenols (Cinnamon extract, Cranberry extract, Crocus sativus L. extract, Curcumin, Garlic extract, Ginger extract, Hesperidin, Olive oil, Pomegranate extract, Puerarin, Quercetin, Resveratrol, Sesamin, Tea polyphenols, Total glucosides of paeony) in this study, there are many other polyphenols in the world. The reason why this study did not include other polyphenols is that this study is about the systematic review and meta-analysis of RCTs, so these animal experiments were not included. In order to further provide future researchers with reference information for dietary polyphenols in the treatment of RA, we summarized dietary polyphenols with potential medicinal value. Dietary polyphenols derived from dietary fruits, vegetables and natural herbs for anti-rheumatic activity in the treatment of RA include Stilbenes, Phenolic acids, Flavonoids, etc. (307-310). Among them, phenolic acids such as benzoic acid and cinnamic acid (311). Hydroxybenzoic acids (including gallotannins and ellagitannins) also exhibit anti-RA effects (312). Natural flavonoids also exhibit strong anti-RA activity, such as anthocyanins, flavanols (catechins, epigallocatechin gallate), flavonoids (luteolin, apigenin), flavanones (naringenin), flavonols (quercetin and kaempferol), isoflavones (genistein, daidzein and glycitein) (313, 314). Some herbs have also been shown to exhibit prominent anti-RA activity, including Adhatoda vasica Nees (pyrroloquinazoline), Ajuga bracteosa wall (Withaferin-A, Ajugarin-I), Aconitum carmechaeli Debeaux (aconitine, benzoylmethasone), Barleria prionitis (triterpenoids), pine (pinitol) (315). These polyphenols play an anti-RA role mainly by regulating immune-inflammation-related signaling pathways, antioxidation-related signaling pathways and triggering synoviocyte/ pathogenic immune cell programmed death (apoptosis, pyroptosis, ferroptosis) (316)(317)(318)(319). At present, the most studied pathways are immune inflammation-related signaling pathways to reduce bone and joint damage, such as NFKB signaling pathways, TLR signaling pathways, TNF signaling pathways, etc. (320,321). The antioxidantrelated signaling pathways regulated by these polyphenols mainly include SIRT1 signaling pathway, Keap1-Nrf2-ARE pathway, NOX4/ROS signaling pathway, HO-1 signaling pathway and iNOS signaling pathway (320,322,323). These dietary polyphenols can also regulate programmed cell death signaling pathways to treat RA, such as MAPK pathway or PI3K/Akt signaling pathway and epigeneticsrelated regulatory signaling pathways (316,324,325).

Strength and limitations
The strengths of this research are that compared with previous studies, this study included the results of a meta-analysis and included more RCTs (48 RCTs were included in this study and involved 15 dietary polyphenols: Cinnamon extract, Cranberry extract, Crocus sativus L. extract, Curcumin, Garlic extract, Ginger extract, Hesperidin, Olive oil, Pomegranate extract, Puerarin, Quercetin, Resveratrol, Sesamin, Tea polyphenols, Total glucosides of paeony). Sensitivity analysis, publication bias assessment, meta-regression, and quality of evidence ratings were also performed in this meta-analysis.
The limitations of this research are that: (1) There is obvious heterogeneity in outcomes. According to meta-regression, part of the heterogeneity of total glucosides of paeony and curcumin may come from drug dosage and intervention time. The remaining heterogeneity may be due to the selection of the population, the choice of dietary polyphenol preparations for treatment of dietary polyphenols, and information biases during data collection. (2) Although a total of 48 RCTs were included, except for total glucosides of paeony, each type of dietary polyphenol supplements included no more than 10 RCTs, and the number of participants in each RCT was mostly less than 100. (3) The RCT languages included in this study were only Chinese and English, and no RCTs in other languages were found, which may have an impact on the results. (4) The follow-up time of the RCTs ranged from 8 to 48 weeks, and there was no observation for more than 3 years and before 8 weeks, which may affect the generality of the results. (5) The RCTs in this research have a high risk of bias. The authors of some RCTs are funded or employed by drug manufacturers, which may introduce bias.

Inspiration for the future
Through the systematic review and meta-analysis of this study, it can be found that the research on the efficacy and mechanism of various dietary polyphenols in relieving RA has made significant progress. Among them, curcumin and total glucosides of paeony have improved the treatment of rheumatoid arthritis. Symptoms, inhibition of inflammation and other outcome indicators. In general, dietary polyphenols can be used as an effective antiinflammatory drug in the adjuvant treatment of RA, and can be used for daily dietary supplementation with relatively good safety.
However, there are still some knowledge gaps in current clinical research and basic research, which need to be addressed in future research. In terms of basic research, most previous studies have cultured a single phenolic compound or a mixture of multiple phenolic compounds with synovial cells, lacking pharmacokinetic studies targeting dietary polyphenols in animal studies. This situation is even more lacking in clinical studies, and few clinical studies have explored the pharmacokinetics of dietary polyphenols in RA patients (such as, absorption and metabolism of polyphenols, and interactions with gut microbiota). For example, in previous studies, most studies using animal or cell culture models focused on the biological activity of polyphenols themselves. Moreover, few studies have focused on the microbial metabolism of polyphenols. In particular, the current study (326)(327)(328) showed that the a-diversity and b-diversity of gut microbiota in RA patients were significantly lower than those in healthy people. Among them, Bacteroidetes and Shiga toxoidproducing Escherichia coli were more abundant in RA patients, in contrast, Lactobacillus, Ravivibacterium isoprawizia, Enterobacter and Stinkbacter were less abundant in the RA group. More studies have found that the intestinal flora can affect the development of RA by regulating immune molecules (including immune cells (such as regulatory B cells and CD4 T helper T cells, etc.), immune factors, etc.) (329)(330)(331)(332). Therefore, research on the interaction between dietary polyphenols and gut microbiota of RA disease is the main direction of future basic and clinical research. In addition, most dietary polyphenols are degraded by gut flora in the colon due to poor polyphenol absorption. The biological activity of dietary polyphenols may be partially attributed to microbial metabolites. The exact role or biological activity of microbial metabolites should be focused on in future studies.

Prospects
Dietary polyphenols are a group of biologically active phytochemicals with a wide range of sources, diverse structures, low toxicity, and good biological activity. In terms of curative effect in the treatment of RA, it is mainly attributed to the immunoinflammation modulating activity of dietary polyphenols, which provides a new option for the comprehensive management strategy of RA, and dietary polyphenols also alleviate the adverse reactions of DMARDs. Since these dietary polyphenols are ubiquitously present in the diet, medication adherence may be better in RA patients. Meanwhile, studies have found that some dietary polyphenols can reduce the side effects of glucocorticoids (such as bone destruction, drug resistance, etc.) (333)(334)(335)(336)(337)(338). Therefore, the current preliminary preclinical data on dietary polyphenols indicate that polyphenols have great potential in the treatment of RA, and the dietary types that have been carried out RCTs are mainly Cinnamon extract, Cranberry extract, Crocus sativus L. extract, Curcumin, Garlic extract, Ginger extract, Hesperidin, Olive oil, Pomegranate extract, Puerarin, Quercetin, Resveratrol, Sesamin, Tea polyphenols, Total glucosides of paeony. In addition, there are many dietary polyphenols such as rutin, chlorogenic acid, anthocyanins, luteolin, lignans, etc., which need extensive clinical research in the future to provide more clinical evidence for the future clinical treatment of RA. Dietary polyphenols have multi-target effects on signaling pathways in RA, so it is important to further explore the pharmacological mechanism of these dietary polyphenols, which will also provide the molecular structure core of lead compounds or new drugs for RA drug development. Based on this systematic review and meta-analysis, we look forward to clinicians and patients using dietary polyphenols as adjunctive therapy or in combination with other current DMARDs in a comprehensive management strategy for RA. In addition, since some polyphenols are mainly derived from herbs, it is necessary in clinical practice to study the pharmacokinetic results of the interaction of herbs as a multicomponent drug (not only polyphenols) with DMARDs to prevent potential adverse drug reactions. Some polyphenol active substances have been formulated into new delivery systems to improve solubility, permeability and enhance the therapeutic effect in preclinical studies due to problems such as absorption. In the future, it is necessary to explore the efficacy and safety of this new type of administration.

Conclusion
Although the number of RCTs on dietary polyphenols is limited, existing evidence shows their potential benefits, mainly increasing DAS28, reducing CRP and ESR, and improving oxidative stress, etc. However, given the small number of patients recruited, the study designs varied widely between RCTs and the characteristics of RA patients varied; it is difficult to immediately extrapolate these results to RA patients in general. More RCTs are needed in the future to determine the efficacy and safety of dietary polyphenols.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Author contributions
ZL, WaX, QH and WeX are responsible for the study concept and design. ZL, WX, QH, WeX, HW, HL, HG, YC, MY, XY, LZ, KY, YD, ZH are responsible for the data collection, data analysis and interpretation; ZL, WaX and KY drafted the paper; ZH and WeX supervised the study; all authors participated in the analysis and interpretation of data and approved the final paper.

Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material
The