Article Text

Higher serum levels of short-chain fatty acids are associated with non-progression to arthritis in individuals at increased risk of RA
  1. Klara Martinsson1,
  2. Kerstin Dürholz2,3,
  3. Georg Schett2,3,
  4. Mario M Zaiss2,3,
  5. Alf Kastbom1
  1. 1 Department of Biomedical and Clinical Sciences, Linköping University, Linköping, Sweden
  2. 2 Department of Internal Medicine 3, Rheumatology and Immunology, Friedrich-Alexander-University Erlangen-Nürnberg (FAU) and Universitätsklinikum Erlangen, Erlangen, Germany
  3. 3 Deutsches Zentrum Immuntherapie (DZI), Friedrich-Alexander-University Erlangen-Nürnberg (FAU) and Universitätsklinikum Erlangen, Erlangen, Germany
  1. Correspondence to Dr Mario M Zaiss, FAU, Erlangen 91054, Germany; mario.zaiss{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Transition from the autoimmune to the clinical phase of rheumatoid arthritis (RA) is a critical step that is yet insufficiently understood. Identification of factors that facilitate the progression from this prodromal RA at-risk state to clinical RA may open new possibilities for preventive interventions. In this context, nutritional factors may be critical. Short-chain fatty acids (SCFAs) are intestinal microbial metabolites that result from nutritional fibre digestion and exert immune regulatory properties.1 SCFAs have shown to effectively inhibit the onset of experimental arthritis.2 Furthermore, serum butyrate levels decrease shortly before the onset of arthritis.2 Whether SCFA levels may play a role in the transition from the autoimmune to the clinical phase of RA in humans, however, has not been studied to date.

To address this concept, we measured serum SCFA levels in a prospective cohort of 82 individuals with an increased risk to develop RA.3 At inclusion, these individuals were positive for anti-citrullinated protein antibodies (ACPA) and had musculoskeletal pain but no clinical signs of arthritis (joint swelling). Baseline characteristics are shown in online supplemental table 1. Following a median follow-up of 72 months, 39 patients (48%) had developed clinical arthritis after a median of 6 months. Baseline serum samples were analysed for SCFA concentrations as previously described.4 At-risk individuals not progressing to arthritis had significantly higher mean baseline serum levels of total SCFA (ie, the sum of acetate, butyrate, propionate or pentanoate), butyrate and acetate as compared by t-test to individuals who progressed to arthritis (figure 1). In contrast, levels of propionate and pentanoate did not significantly differ (figure 1). Univariable Cox regression analyses revealed significant association between lower total SCFA levels and progression to arthritis (p=0.029), while for the individual SCFA, we found significant associations concerning butyrate (p=0.038) and acetate (p=0.039) levels, but not regarding pentanoate or propionate (online supplemental table 2). Statistical significance remained after adjusting for age, sex, symptom duration, rheumatoid factor status, ACPA levels and CRP levels (total SCFA p=0.030; butyrate p=0.009 and acetate p=0.045, online supplemental table 2).

Figure 1

Baseline serum samples from rheumatoid arthritis at-risk individuals (ACPA+; musculoskeletal pain+) progressing (n=39) or not progressing (n=43) to arthritis in a prospective observational cohort study3 were analysed for (A) acetate, (B) butyrate, (C) pentanoate and (D) propionate levels. Bars represent means and error bars represent SD. ACPA, anti-citrullinated protein antibodies.

Butyrate levels inversely correlated with serum IgA-ACPA levels (r=−0.23, p=0.039), but not with IgG-ACPA or IgM-ACPA. No other SCFAs were significantly correlated with any ACPA subtype.

These data suggest that SCFA, in particular butyrate and acetate, influences the risk for the transition from the autoimmune to the clinical phase of RA. Although most p values would not remain significant after correction for multiple testing, the data are in line with previous findings in animal models2 and thus confirm our prespecified hypothesis. As SCFAs are produced by intestinal microbiota on fermentation of dietary fibres, our findings strengthen the concept that nutritional factors could influence the onset of RA. SCFAs are critical for the barrier function of the intestinal epithelium and thereby influences the migration of cells from the gut to the joints.2 Increasing SCFA levels by direct supplementation, fiber-rich diet or faecal transplantation to restore early dysbiosis thus represent potential strategies to inhibit the onset of arthritis.4–6 In this context, high-fibre diet has already shown to increase SCFA levels and decrease inflammatory burden in patients with established RA4 but has not been applied in a preventive setting. These data suggest that a state of high SCFA concentrations, which can be reached by dietary interventions such as high-fibre diet, may go along with a lower risk to progress to clinical arthritis in individuals with a high risk to develop RA.

Ethics statements

Patient consent for publication

Ethics approval

We declare that this study was approved by Swedish Ethical Review Authority (2020-06162) and was conducted in accordance with the ethical standards set forth in the Declaration of Helsinki and its subsequent amendments. All study subjects provided written informed consent before inclusion in the TIRx cohort.


We wish to thank Dr Jörg Hoffmann from the Bioanalytics Group at Department of Biology, FAU, Erlangen, Germany, for the analysis of the samples. We are also grateful to the TIRx patients and collaborators in Linköping, Sweden.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Handling editor Josef S Smolen

  • Contributors All authors had access to the data and a role in writing the manuscript. KMKM, MMZ, GS and AK contributed to the study conception and design. AK was responsible for patient recruitment and characterisation. Sample preparation was done by KD. Data was collected by KD and MMZ. Statistical analyses were performed by KMKM. MMZ and AK wrote the first manuscript draft and all authors commented on posterior versions and approved the final manuscript.

  • Funding AK is funded by the King Gustaf V’s 80-year foundation, the Swedish Rheumatism association, and ALF grants from Region Östergötland. KD, GS and MMZ are funded by the Deutsche Forschungsgemeinschaft (DFG-FOR2886 project A1 and CRC1181 project B07), the Bundesministerium für Bildung und Forschung (BMBF; project MASCARA), the H2020 GA 810316-4D-Nanoscope ERC Synergy Project, the IMI funded projects RTCure and HIPPOCRATES, the Emerging Fields Initiative MIRACLE of the Friedrich-Alexander-Universität Erlangen-Nürnberg.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.