Exploring gut microbiota's role in rheumatic valve disease: insights from a Mendelian randomization study and mediation analysis

Abstract

Background: Investigating the relationship between gut microbiota and Rheumatic Valve Disease (RVD) is crucial for understanding the disease's etiology and developing effective interventions. Our study adopts a novel approach to examine the potential causal connections between these factors.

Methods: Utilizing a two-sample Mendelian Randomization (MR) framework, we incorporated a multi-variable MR (MVMR) strategy to assess the mediatory mechanisms involved. This approach involved analyzing data from the MiBioGen consortium for gut microbiota and the FinnGen for RVD, among other sources. Instrumental variables (IVs) were carefully selected based on rigorous MR principles, and statistical analysis was conducted using bidirectional two-sample MR, such as inverse variance-weighted (IVW), weighted median, MR-Egger regression and MR Steiger Test methods. The MR-PRESSO strategy was employed for outlier detection, and MVMR was used to untangle the complex relationships between multiple microbiota and RVD.

Results: Our analysis highlighted several gut microbiota classes and families with potential protective effects against RVD, including Lentisphaerae, Alphaproteobacteria, and Streptococcaceae. In contrast, certain genera, such as Eubacterium eligens and Odoribacter, were identified as potential risk factors. The MVMR analysis revealed significant mediation effects of various immune cell traits and biomarkers, such as CD4^-CD8^- T cells, CD3 on Terminally Differentiated CD8+ T cell and Pentraxin-related protein PTX, elucidating the complex pathways linking gut microbiota to RVD.

Conclusion: This study underscores the intricate and potentially causal relationship between gut microbiota and RVD, mediated through a range of immune and hormonal factors. The use of MVMR in our methodological approach provides a more comprehensive understanding of these interactions, highlighting the gut microbiota's potential as therapeutic targets in RVD management. Our findings pave the way for further research to explore these complex relationships and develop targeted interventions for RVD.

Keywords: Mendelian randomization; estradiol; gut microbiota; immune cell; mediation analysis; rheumatic valve disease.

Figure 1

Mendelian randomization flowchart.

Figure 2

Mendelian randomization core assumption. This Figure represent the MR and mediation assumption, including three core assumption: Independence assumption: SNPs should be robustly associated with the exposure or mediator without confounding factors; SNPs should have a strong link with the exposure or mediator; Exclusion restriction assumption: SNPs should affect the outcome only through the exposure (mediator).

Figure 3

Mendelian randomization analysis between gut microbiota and the risk of RVD. This forest plot represents a Mendelian randomization analysis exploring the association between different classifications of gut microbiota and the risk of Rheumatic Valve Disease (RVD). Each point indicates the Odds Ratio (OR), showing the strength and direction of the association. Horizontal lines represent Confidence Intervals, providing a range for the true OR. Points to the left of the vertical line (OR < 1) suggest a protective effect against RVD, while points to the right (OR > 1) indicate a potential risk. P-value denote the statistical significance of each association.

Figure 4

Mendelian randomization analysis between RVD gut microbiota. This forest plot illustrates a statistical analysis investigating the relationship between Rheumatic Valve Disease (RVD) and gut microbiota. Each horizontal line represents a different microbial classification, with the central dot indicating the beta coefficient (effect size). The span of each line shows the 95% confidence interval, indicating the range within which the true effect size is likely to fall. A beta coefficient to the right of the zero line (positive value) suggests a potential increase microbiota abundance with RVD, while a coefficient to the left (negative value) indicates a potential decrease microbiota abundance with RVD. The annotations to the right of each line provide the beta coefficient and P-value, with the P-value indicating the statistical significance of the association.

Figure 5

Mediation analysis between Mediator and RVD. This forest plot illustrates a statistical analysis investigating the relationship between various immunological markers and hormone levels with the risk of RVD. Each horizontal line represents the odds ratio (OR) for RVD risk. An OR below 1 suggests a potential protective effect, while an OR above 1 indicates a potential increased risk. The span of each line shows the 95% confidence interval, indicating the range within which the true effect size is likely to fall. The annotations above each line provide the OR and P-value, with the P-value indicating the statistical significance of the association.

Figure 6

Mediation analysis between gut microbiota and Mediator. This forest plot illustrates a statistical analysis investigating the relationship between gut microbiota and mediator. Each horizontal line represents the effect of microbial classification on mediator, with the central dot indicating the beta coefficient (effect size). The span of each line shows the 95% confidence interval, indicating the range within which the true effect size is likely to fall. A beta coefficient to the right of the zero line (positive value) suggests a potential increase in RVD risk with that mediator, while a coefficient to the left (negative value) indicates a potential decrease in risk. The annotations above each line provide the beta coefficient and P-value, with the P-value indicating the statistical significance of the association.