- Rheumatoid arthritis (RA) is a chronic autoimmune inflammatory arthritis causing symmetric joint pain, morning stiffness, synovitis, and progressive joint/bone erosion, affecting roughly 0.5–1% of people; the strongest genetic risk is the HLA-DRB1 "shared epitope," and diagnosis is supported by anti-citrullinated protein antibodies (ACPA) and rheumatoid factor (RF) (Alghamdi/ACPA review, PMC).
- DMARDs (methotrexate first-line), biologics (TNF and IL-6 inhibitors), and JAK inhibitors are the disease-modifying standard of care, following EULAR and ACR treat-to-target guidelines — this is Strong evidence. Supplements below are adjuncts only, never replacements.
- Omega-3 fish oil has the most robust independent human-trial evidence among supplements (Moderate–Strong): ~3 g/day EPA+DHA for 3 months reduced tender joint count by RD −2.9 and morning stiffness by 25.9 minutes, with an NSAID-sparing effect (Fortin et al. meta-analysis).
- Curcumin shows Moderate evidence from a 6-RCT meta-analysis (DAS28 MD −1.20) but trials are small and heterogeneous (Kou et al. 2023). Ginger, Mediterranean diet, vitamin D, and probiotics are Weak–Moderate with mixed or surrogate-heavy results.
- Boswellia and devil's claw have no adequate RA-specific RCT (Insufficient for RA) — their supporting data is from osteoarthritis, and the leading boswellia trial (5-Loxin) is funded by manufacturer Laila Impex with employee co-authors, a material conflict of interest.
- Combined bleeding risk: stacking fish oil + curcumin + ginger with NSAIDs or anticoagulants may additively raise bleeding risk — the most clinically relevant supplement-interaction theme for RA patients.
- Evidence grade: Strong (drugs) / Moderate (fish oil, curcumin)
Health Concern Deep Dive · Autoimmune / Joint Health
Table of contents
- Evidence summary table
- What rheumatoid arthritis is
- How it works
- Conventional treatment
- Supplement, nutrition, and lifestyle evidence
- What works and what does not
- Risks and side effects
- All interactions
- Who should avoid it
- Dosage and how to take
- Animal and in-vitro evidence
- Independent funding notes
- Frequently asked questions
- Sources and funding notes
Evidence summary table
| Claim | Evidence | Source | Funding / conflict traced | Strength |
|---|---|---|---|---|
| DMARDs/biologics are the disease-modifying standard of care | EULAR/ACR treat-to-target guidelines recommend methotrexate first-line, escalating to TNF/IL-6/JAK inhibitors for inadequate responders; supported by decades of replicated RCT evidence. | EULAR 2019 update | Independent professional-society guideline synthesizing many trials. | Strong |
| Omega-3 fish oil reduces tender joints and morning stiffness | Validation meta-analysis / mega-analysis: tender joint count RD −2.9 (95% CI −3.8 to −2.1); morning stiffness −25.9 min (−44.3 to −7.5); NSAID-sparing. | Fortin et al.; Kostoglou-Athanassiou et al. 2020 | Academic; independent meta-analysis, consistent across trials. | Moderate–Strong |
| Curcumin improves disease activity markers | 6-RCT meta-analysis (539 patients): DAS28 MD −1.20; ESR −29.47; SJC −5.33; TJC −6.33 vs control. | Kou et al., Front Immunol 2023 | Academic (China); independent meta-analysis, but individual trials are small and heterogeneous. | Moderate |
| Ginger reduces RA disease activity | Single RCT, n=70: significant reduction in disease activity score vs placebo; favorable FoxP3/T-bet/RORγt gene-expression shifts. | Aryaeian et al., Gene 2019 | Academic (Iran); independent, single small RCT. | Weak–Moderate |
| Mediterranean diet improves RA outcomes | MADEIRA RCT plus observational data: diet + activity nudges improved clinical/inflammatory markers vs usual care over 12+ weeks. | Papandreou et al., Nutrients 2023 | Academic (Greece); independent; small RCT plus observational evidence. | Weak–Moderate |
| Vitamin D affects disease activity | Dose-response meta-analysis, 11 RCTs/3,049 patients: no effect on CRP/ESR/HAQ, but DAS28-CRP −0.58 and DAS28-ESR −0.58; high heterogeneity. | Al-Saoodi et al., Nutr Rev 2024 | Academic (University of Copenhagen); independent; high heterogeneity limits confidence. | Weak–Moderate |
| Probiotics reduce inflammation markers | Meta-analysis, 8 RCTs/344 patients: CRP MD −3.04 mg/L, but no DAS28 difference after excluding high-risk-of-bias trials (SMD −0.54, p=0.45). | Sanchez et al., Nutrients 2022 | Academic (France); independent; authors do not recommend for disease control. | Weak |
| Boswellia treats RA | No adequate RA RCT exists; the strongest boswellia trial (5-Loxin) is an osteoarthritis trial, not RA. | Sengupta et al. 2008; NCCIH | 5-Loxin trial funded by Laila Impex (manufacturer); several authors are Laila Impex employees — conflicted. | Insufficient for RA |
| Devil's claw treats RA | Evidence base is for osteoarthritis and low back pain, not RA; no convincing RA-specific RCT. | See osteoarthritis/LBP literature | Mixed; not RA-specific. | Insufficient for RA |
| Type II collagen modifies RA | Older RCTs with mixed results; not adopted into clinical guidelines. | Older oral-tolerance RCT literature | Mixed; inconsistent replication. | Contested |
What rheumatoid arthritis is
Rheumatoid arthritis (RA) is a chronic autoimmune inflammatory arthritis that causes symmetric joint pain, prolonged morning stiffness, synovitis (inflammation of the joint lining), and progressive joint and bone erosion if untreated (Alghamdi/ACPA review, PMC). It typically affects the small joints of the hands and feet first, though it can involve any synovial joint, and its hallmark is symmetry — both wrists or both sets of knuckles rather than one isolated joint. Prevalence is estimated at roughly 0.5–1% of the population, making it one of the more common autoimmune diseases (Alghamdi/ACPA review, PMC). The strongest known genetic risk factor is the HLA-DRB1 "shared epitope" — a group of related HLA-DRB1 alleles that share a similar amino acid sequence in the antigen-binding groove and are strongly associated with RA susceptibility and severity (Alghamdi/ACPA review, PMC).How it works
RA autoimmunity is characterized by two key antibody families: anti-citrullinated protein antibodies (ACPA) and rheumatoid factor (RF). Citrullination is a post-translational modification in which the enzyme peptidylarginine deiminase converts the amino acid arginine into citrulline within certain proteins, creating new "neoantigens" that the immune system can mistakenly target (ACPA pathogenesis review, PMC). This citrullination-driven immune response is thought to be a central driver of synovitis and the bone erosion seen in established RA. Genetic predisposition alone is not sufficient to cause RA. Smoking and infections are key environmental triggers that appear to act on genetically susceptible individuals — particularly those carrying the HLA-DRB1 shared epitope — tipping the immune system toward autoimmunity (ACPA pathogenesis review, PMC). This gene-environment interaction model helps explain why RA clusters in some families and some smokers but does not develop in everyone with the shared epitope.Conventional treatment
The evidence-based standard of care for RA follows a treat-to-target strategy: start disease-modifying therapy early, measure disease activity regularly (commonly with the DAS28 composite score), and escalate treatment until low disease activity or remission is reached. Methotrexate is the first-line conventional synthetic DMARD, with biologics — TNF inhibitors and IL-6 inhibitors — and targeted synthetic JAK inhibitors used for patients with an inadequate response, per the EULAR and ACR guidelines (EULAR 2019 update). NSAIDs are used for symptom control (pain and stiffness) but do not modify disease progression, and corticosteroids are used as short-term "bridging" therapy while DMARDs take effect, because of their own long-term side-effect burden. These pharmaceutical interventions have the strongest, most replicated randomized-controlled-trial evidence for disease control available for RA, and are the standard of care against which everything else — including every supplement discussed below — must be judged as an adjunct, not a replacement (EULAR 2019 update).Supplement, nutrition, and lifestyle evidence
| Intervention | Evidence type (human) | Effect size / finding | Dose / duration | Grade |
|---|---|---|---|---|
| Omega-3 fish oil | Meta-analysis + mega-analysis of RCTs; narrative review. | 3 months: tender joint count RD −2.9 (95% CI −3.8 to −2.1, p=0.001); morning stiffness −25.9 min (−44.3 to −7.5, p<0.01). Modest benefit plus NSAID-sparing effect. | ~3 g/day EPA+DHA, ≥3 months | Moderate–Strong |
| Curcumin / turmeric | Systematic review + meta-analysis, 6 RCTs, 539 patients. | DAS28 MD −1.20 (95% CI −1.85 to −0.55, p=0.0003); ESR −29.47; SJC −5.33; TJC −6.33 vs control. Some RCTs report comparability to NSAIDs. | ~500 mg bioavailable curcumin 2×/day, 8–12 wk | Moderate |
| Ginger | Randomized double-blind placebo-controlled trial, n=70. | Significant reduction in disease activity score vs placebo; favorable shifts in FoxP3/T-bet/RORγt gene expression. | 1,500 mg/day × 12 wk | Weak–Moderate |
| Mediterranean diet | MADEIRA randomized controlled trial in women with RA, plus observational evidence. | Diet + physical-activity nudges vs usual care showed clinical and inflammatory improvements. | 12+ weeks | Weak–Moderate |
| Vitamin D | Dose-response meta-analysis, 11 RCTs, 3,049 patients. | No significant effect on CRP, ESR, or HAQ; but reduced pain-VAS (WMD −1.30), DAS28-CRP −0.58, DAS28-ESR −0.58; highly heterogeneous. | Varies; >100 µg/day more effective on CRP | Weak–Moderate |
| Probiotics | Systematic review + meta-analysis, 8 RA RCTs, 344 patients. | CRP MD −3.04 mg/L (p<0.001); but after excluding high-risk-of-bias trials, no DAS28 difference (SMD −0.54, p=0.45, I²=93%). | 8 wk–1 yr | Weak |
| GLA (evening primrose / borage oil) | Older, small RCTs. | Some reduction in tender/swollen joints in older trials; limited replication. | ~1.4–2.8 g GLA/day | Weak |
| Boswellia (frankincense) | No adequate RA RCT. Strongest trial (5-Loxin) is osteoarthritis, not RA. | 5-Loxin improved OA (not RA) knee pain/function at 100–250 mg/day over 90 days. NCCIH: topical boswellia for RA "not enough evidence." | — | Insufficient for RA |
| Devil's claw (Harpagophytum) | Evidence is largely for osteoarthritis / low back pain, not RA. | No convincing RA-specific RCT. | — | Insufficient for RA |
| Type II collagen (oral tolerance) | Older RCTs, mixed results. | Inconsistent; not adopted into guidelines. | — | Weak / Contested |
Omega-3 fish oil: the best-supported supplement
Among all supplements studied for RA, omega-3 fish oil has the most robust independent human-trial evidence. A validation meta-analysis and mega-analysis of RCTs found that roughly 3 g/day of EPA+DHA for 3 months produced a clinically noticeable drop in tender joint count (RD −2.9) and shortened morning stiffness by nearly 26 minutes, alongside allowing some patients to reduce their NSAID use (Fortin et al., Kostoglou-Athanassiou et al. 2020). This is an academic, independent meta-analysis with consistent findings across the included trials.Curcumin: a positive but heterogeneous signal
A 2023 systematic review and meta-analysis pooling 6 RCTs and 539 patients found that curcumin supplementation reduced DAS28 disease-activity scores by a mean difference of −1.20, alongside reductions in ESR, swollen joint count, and tender joint count compared with control (Kou et al., Front Immunol 2023). Some of the individual trials even reported effects comparable to NSAIDs. This is an independent academic meta-analysis from China, but the underlying trials are small and heterogeneous, which is why the grade is Moderate rather than Strong.Ginger, Mediterranean diet, vitamin D, and probiotics
Ginger has one randomized, double-blind, placebo-controlled trial of 70 patients showing a significant reduction in disease activity score at 1,500 mg/day over 12 weeks, along with favorable shifts in T-cell-related gene expression (Aryaeian et al., Gene 2019). This is a single small independent RCT, so while promising, it needs replication. The MADEIRA randomized controlled trial tested a Mediterranean diet combined with physical-activity encouragement in women with RA and found clinical and inflammatory improvements versus usual care over 12 or more weeks, consistent with broader observational evidence on diet and inflammation (Papandreou et al., Nutrients 2023). Vitamin D evidence is mixed: an 11-RCT, 3,049-patient dose-response meta-analysis found no significant effect on CRP, ESR, or the Health Assessment Questionnaire (HAQ), but did find modest reductions in pain and DAS28 scores, with very high heterogeneity across trials limiting confidence (Al-Saoodi et al., Nutr Rev 2024). Probiotics reduced CRP by a mean difference of −3.04 mg/L across 8 RCTs and 344 patients, but once trials at high risk of bias were excluded, there was no significant difference in DAS28 disease activity — and the review authors themselves do not recommend probiotics for RA disease control (Sanchez et al., Nutrients 2022).Boswellia and devil's claw: insufficient evidence for RA
Type II collagen: contested
Oral type II collagen, based on an "oral tolerance" immunological theory, has been tested in older RCTs with mixed and inconsistent results. It has not been adopted into EULAR, ACR, or other major RA treatment guidelines, and is best described as Contested rather than supported.What works and what does not
| Intervention | Verdict | Evidence | Key caveat |
|---|---|---|---|
| DMARDs (methotrexate) / biologics / JAK inhibitors | WORKS — standard of care | EULAR/ACR treat-to-target guidelines built on decades of replicated RCTs (EULAR 2019 update). | Requires clinician monitoring for toxicity and infection risk. |
| Omega-3 fish oil | WORKS as an adjunct | Meta-analysis: fewer tender joints, shorter morning stiffness, NSAID-sparing (Fortin et al.). | Modest effect size; not disease-modifying on its own. |
| Curcumin | LIKELY HELPS, but evidence is heterogeneous | 6-RCT meta-analysis found DAS28 improvement (Kou et al. 2023). | Small individual trials; bioavailable formulations vary widely. |
| Ginger | PRELIMINARY positive signal | One small RCT showed disease-activity reduction (Aryaeian et al. 2019). | Needs replication in larger trials. |
| Mediterranean diet | LIKELY HELPFUL as lifestyle adjunct | MADEIRA RCT plus observational data (Papandreou et al. 2023). | Small trial; general cardiometabolic benefits likely contribute. |
| Vitamin D | MIXED | No CRP/ESR/HAQ benefit, but modest DAS28/pain improvement across 11 RCTs (Al-Saoodi et al. 2024). | Very high heterogeneity between trials. |
| Probiotics | MIXED — CRP only | CRP reduced, but no reliable DAS28 benefit after bias correction (Sanchez et al. 2022). | Authors do not recommend for disease-activity control. |
| Boswellia | INSUFFICIENT for RA | No adequate RA RCT; best trial is OA and manufacturer-conflicted (Sengupta et al. 2008). | Do not rely on OA-borrowed data for RA claims. |
| Devil's claw | INSUFFICIENT for RA | Evidence is from OA/low back pain populations, not RA. | No RA-specific efficacy trial identified. |
| Type II collagen | CONTESTED | Mixed older RCTs; not guideline-adopted. | Do not substitute for DMARDs. |
Risks and side effects
| Ingredient | Common side effects | Rare / serious | Source |
|---|---|---|---|
| Omega-3 fish oil | Fishy aftertaste, GI upset, belching | — | Fortin et al. |
| Curcumin / turmeric | GI upset, diarrhea | Rare hepatotoxicity reports, especially with piperine-boosted or high-dose products | Kou et al. 2023; safety-signal literature |
| Ginger | Mild GI upset, heartburn | — | Aryaeian et al. 2019 |
| Boswellia | GI upset, nausea, diarrhea | — | NCCIH (reported safe ≤1,000 mg/day up to 6 months in trials) |
| Vitamin D | Usually none at standard doses | Hypercalcemia with excessive dosing | Al-Saoodi et al. 2024 |
| Probiotics | Bloating, gas | Rare bacteremia in immunocompromised patients | Sanchez et al. 2022 |
All interactions
| Interacts with | Ingredient(s) | Severity | Mechanism | Action |
|---|---|---|---|---|
| Anticoagulants / antiplatelets (warfarin, DOACs, aspirin, clopidogrel) | Omega-3 fish oil, curcumin, ginger | Caution / monitor | Theoretical additive bleeding risk; large independent data suggest clinically significant bleeding risk from fish oil alone is low even at higher doses, but the combined-stack risk with curcumin and ginger is the key concern. | Discuss with a prescriber before combining any two or more of these with a blood thinner or NSAID; monitor for bruising/bleeding. |
| NSAIDs (ibuprofen, naproxen, diclofenac) | Omega-3 fish oil, curcumin, ginger | Caution | Additive bleeding/GI risk when combined with regular NSAID use common in RA symptom control. | Monitor for GI bleeding signs; do not exceed labeled NSAID doses. |
| Drug-metabolizing enzymes and iron absorption | Curcumin | Caution | May affect cytochrome-mediated drug metabolism and reduce iron absorption. | Space from iron supplements; flag to prescriber if on narrow-therapeutic-index medications. |
| Thiazide diuretics | Vitamin D | Monitor | Additive hypercalcemia risk when combined with thiazides. | Monitor calcium levels, especially with high-dose vitamin D. |
| Immunosuppressants (including DMARDs/biologics used for RA itself) | Probiotics | Caution in severe immunosuppression | Rare bacteremia risk in immunocompromised patients taking live probiotic organisms. | Discuss with rheumatologist before starting probiotics on biologics or high-dose immunosuppression. |
| Antidiabetics | Ginger | Monitor | Possible additive glucose-lowering effect. | Monitor blood glucose if combining with insulin or oral antidiabetics. |
Who should avoid it
- Pregnancy: High-dose curcumin, boswellia, and several other botanicals discussed here lack adequate safety data in pregnancy and should generally be avoided unless a clinician approves.
- Gallstones or bile duct obstruction: Curcumin can stimulate bile flow and should be avoided or used cautiously in people with gallstones or biliary obstruction.
- Bleeding disorders or upcoming surgery: Anyone with a bleeding disorder, or who is scheduled for surgery, should avoid combining fish oil, curcumin, and ginger, and should discuss all supplements with their surgical team in advance.
- Immunocompromised patients: People on biologics, high-dose corticosteroids, or other significant immunosuppression should avoid starting probiotics without medical guidance, due to rare reports of bacteremia in immunocompromised hosts.
- Anyone currently untreated or undertreated for RA: No one should use supplements as a substitute for starting or continuing DMARD/biologic therapy; delaying effective treatment risks irreversible joint damage.
Dosage and how to take
| Intervention | Typical studied dose | Duration used in trials | Practical note |
|---|---|---|---|
| Omega-3 fish oil | ~3 g/day combined EPA+DHA | ≥3 months before assessing effect | Take with meals to reduce fishy aftertaste and GI upset; discuss with prescriber if on anticoagulants. |
| Curcumin | ~500 mg bioavailable curcumin, twice daily | 8–12 weeks | Use a bioavailability-enhanced formulation; watch for GI upset; avoid high-dose piperine-boosted products if at hepatotoxicity risk. |
| Ginger | 1,500 mg/day | 12 weeks in the supporting RCT | Single small trial; treat as an adjunct trial, not a guaranteed benefit. |
| Mediterranean diet | Whole-diet pattern plus physical activity encouragement | 12+ weeks in the MADEIRA RCT | Low risk, generally beneficial for cardiometabolic health regardless of RA-specific effect size. |
| Vitamin D | Varies; >100 µg/day showed more effect on CRP in pooled data | Varies across the 11 pooled RCTs | Check baseline vitamin D level with a clinician before high-dose supplementation. |
| Probiotics | Varies by strain and product | 8 weeks to 1 year across pooled trials | Not recommended as a primary disease-activity strategy based on current evidence. |
Animal and in-vitro evidence
This article relies exclusively on independent human-trial evidence for all efficacy and safety conclusions. Animal studies and in-vitro (cell-culture) research on omega-3s, curcumin, ginger, boswellia, and other RA-related supplements exist in the broader literature but are excluded from every conclusion above, consistent with an evidence standard that treats animal and non-human-relevant in-vitro data as insufficient to establish human efficacy or safety. Where mechanistic background is mentioned (for example, citrullination biology in the "How it works" section), it is presented strictly as background context, not as evidence of any treatment effect.Independent funding notes
| Evidence source | Country | Funding / conflicts | Independence rating |
|---|---|---|---|
| Fortin et al., omega-3 validation meta-analysis | Academic, multi-trial synthesis | No industry funding identified; consistent findings across independently conducted trials. | Independent |
| Kou et al., curcumin meta-analysis, Front Immunol 2023 | China | Academic institutional research; independent meta-analysis of 6 RCTs. | Independent |
| Aryaeian et al., ginger RCT, Gene 2019 | Iran | Academic; independent single RCT. | Independent |
| Papandreou et al., MADEIRA RCT, Nutrients 2023 | Greece | Academic; independent. | Independent |
| Al-Saoodi et al., vitamin D meta-analysis, Nutr Rev 2024 | Denmark (University of Copenhagen) | Academic; independent; high heterogeneity flagged by the authors themselves. | Independent |
| Sanchez et al., probiotics meta-analysis, Nutrients 2022 | France | Academic; independent; authors explicitly do not recommend probiotics for disease control based on their own findings. | Independent |
| Sengupta et al., 5-Loxin boswellia trial, Arthritis Res Ther 2008 | India | Funded by Laila Impex, the 5-Loxin manufacturer; several authors are Laila Impex employees. | Conflicted — flagged, not used to support any RA claim |
| EULAR 2019 update | Europe (multinational professional society) | Independent professional-society guideline process synthesizing many independently funded trials. | Independent |
The overall pattern is consistent with the broader independent-evidence literature: the omega-3 and curcumin meta-analyses are independent academic work with no material industry funding detected, while the leading boswellia trial is manufacturer-funded with employee co-authors and is explicitly downgraded and excluded from supporting any RA-specific claim.
Frequently asked questions
Can supplements replace DMARDs or biologics for rheumatoid arthritis?
No. DMARDs (methotrexate first-line) and biologics are the disease-modifying standard of care with Strong RCT evidence, following EULAR/ACR treat-to-target guidelines. Supplements such as omega-3 fish oil and curcumin are adjuncts that may help with symptoms, but none has evidence of preventing the joint erosion that untreated or undertreated RA causes (EULAR 2019 update).
What is the best-supported supplement for RA?
Omega-3 fish oil has the most robust independent human-trial evidence among RA supplements, with a meta-analysis showing reduced tender joint count (RD −2.9) and shorter morning stiffness (−25.9 minutes) at about 3 g/day EPA+DHA for at least 3 months, along with an NSAID-sparing effect (Fortin et al.).
Does curcumin work as well as NSAIDs for RA?
Some individual trials within a 6-RCT meta-analysis reported curcumin effects comparable to NSAIDs, and the pooled analysis found a DAS28 improvement of −1.20. However, the underlying trials are small and heterogeneous, so this should be treated as a Moderate-grade, promising signal rather than an NSAID-equivalence guarantee (Kou et al. 2023).
Is boswellia proven to help rheumatoid arthritis?
No. There is no adequate RA-specific randomized controlled trial for boswellia. Its most-cited supporting trial (5-Loxin) tested osteoarthritis, not RA, and was funded by the manufacturer Laila Impex with employee co-authors — a direct conflict of interest. NCCIH states there is not enough evidence for boswellia in RA (NCCIH).
Is it safe to combine fish oil, curcumin, and ginger?
Combining all three, especially alongside NSAIDs or anticoagulants, may additively increase bleeding risk. This is the most clinically relevant supplement-interaction concern for RA patients and should be discussed with a prescriber, particularly for anyone also using NSAIDs or blood thinners.
Does vitamin D help rheumatoid arthritis?
The evidence is mixed. An 11-RCT, 3,049-patient meta-analysis found no significant effect on CRP, ESR, or HAQ, but did find modest improvements in pain and DAS28 scores, with very high heterogeneity between trials limiting confidence in a specific dose-response recommendation (Al-Saoodi et al. 2024).
Sources and funding notes
- Alghamdi et al. — ACPA pathogenesis and HLA-DRB1 shared epitope review, PMC
- EULAR recommendations for management of rheumatoid arthritis with synthetic and biological DMARDs: 2019 update
- Fortin et al. — Validation meta-analysis of omega-3 fatty acids in rheumatoid arthritis
- Kostoglou-Athanassiou et al. — Omega-3 fatty acids and rheumatoid arthritis, Mediterr J Rheumatol 2020
- Kou et al. — Curcumin for rheumatoid arthritis: systematic review and meta-analysis, Front Immunol 2023
- Aryaeian et al. — Ginger RCT in rheumatoid arthritis, Gene 2019
- Papandreou et al. — MADEIRA Mediterranean diet RCT in RA, Nutrients 2023
- Al-Saoodi et al. — Vitamin D dose-response meta-analysis in RA, Nutr Rev 2024
- Sanchez et al. — Probiotics meta-analysis in RA, Nutrients 2022
- Sengupta et al. — 5-Loxin boswellia trial (osteoarthritis), Arthritis Res Ther 2008
- NCCIH — Boswellia
