Genetic relationship between rheumatoid arthritis and cardiovascular diseases

RA is a chronic autoimmune disease characterized by systemic inflammation and synovial joint destruction. It affects approximately 1% of the global population, with a higher prevalence among women [1]. While RA primarily targets the joints, accumulating evidence suggests that it also exerts systemic effects, impacting various organ systems beyond the musculoskeletal system [2]. Of particular concern is the well-established association between RA and cardiovascular diseases (CVD), which has garnered increasing attention in recent years. Individuals with RA face a significantly elevated risk of developing CVD, including conditions such as atherosclerosis, coronary artery disease, myocardial infarction, stroke and heart failure [3]. This is also increasingly being demonstrated for other autoimmune diseases [4]. Multiple factors contribute to this heightened cardiovascular risk among RA patients, including chronic inflammation, traditional cardiovascular risk factors and the potential side effects of RA medication, especially glucocorticoids [5]. Chronic inflammation, a hallmark of RA, is increasingly recognized as a key driver in the development and progression of atherosclerosis, making RA a unique model for exploring the complex interplay between chronic inflammatory diseases and CVD [5]. There are data for RA showing that achieving a state of low disease activity or remission leads to a cardiovascular risk comparable to that of the general population [6].

Specific SNPs such as rs2476601 (PTPN22), rs1801274 (FCGR2A), rs651007 (IL6R) and rs10774624 (TNFAIP3) are associated with genes regulating inflammation and immune responses, highlighting their role in the genetic mechanisms underlying RA and the cardiovascular implications [7‐10]. This overview underlines the importance of understanding genetic variations to elucidate disease pathways and potential therapeutic targets [8‐10]. To investigate the causal relationship between RA and CVD and to decipher potential genetic mechanisms, researchers have turned to Mendelian randomization (MR) studies. This is a powerful and innovative approach that leverages genetic variants as instrumental variables to estimate causal associations between exposure and outcomes [11, 12]. Unlike traditional observational studies, MR studies are less susceptible to confounding and reverse causality, offering a robust method for assessing causality in complex disease relationships. These genetic instruments help to understand whether RA causally contributes to the development of CVD or if the observed association is confounded by shared risk factors. By employing MR connections between RA-related inflammation, genetic predisposition and the collective impact on CVD are investigated [12].

Advancements in genetics and the availability of large-scale genome-wide association studies (GWAS) datasets have facilitated the application of MR to explore the causal links between RA and CVD. These studies aim to explain whether the systemic inflammation and genetic factors associated with RA play a direct role in the pathogenesis of CVD or if other mediators are involved. Understanding the causal pathways between RA and CVD could have the potential to find treatment strategies for both conditions and improve the cardiovascular outcomes of RA patients [11, 13].

In light of the increasing interest in the connection between RA and CVD, it is imperative to conduct a systematic review to consolidate the findings of MR studies and offer a representation of the current knowledge landscape. This review aims not only to deepen the comprehension of the causal relationship between RA and CVD but also to direct future research.

The systematic review was registered on Open Science Framework (OSF). A comprehensive literature search was conducted using PubMed, Cochrane and Web of Science databases. The search was conducted using the following search terms: “rheumatoid arthritis” AND “Mendelian.” Two reviewers (MA and BW) independently assessed the eligibility of these studies based on predefined inclusion and exclusion criteria. The search was completed as of 15 October 2023. All studies conducted up to that date were considered for the review. The inclusion criteria focused on studies investigating the association between RA and CVD using MR methods. Studies were excluded if they did not meet these criteria, if they were reviews or comments, if they were not available in full text, or if they were not fully published in English. Additionally, the MR studies had to specifically address cardiovascular risk or diseases. The two reviewers identified the studies that met the inclusion criteria and were deemed suitable for the systematic review based on relevance, study design and the use of MR to examine the relationship between RA and CVD. The search process is depicted in the Prisma flow chart (Fig. 1). The selected studies are shown here:

To evaluate the quality of the selected studies an adapted critical appraisal checklist for MR studies in CVD and RA research. The original checklist was introduced in the paper “Reading Mendelian randomization studies: a guide, glossary, and checklist for clinicians” [11]. This checklist was partly tailored to assess the methodology, outcomes and implications of each MR study within the framework of RA as a potential risk factor for CVD. The authors first ensured that the checklist questions directly pertained to the central research topic, which focused on the connection between RA and CVD. Next, adjustments were made to the questions to enhance the specificity, aligning them with the study’s objectives. To facilitate organization and accessibility, the questions were grouped into categories with appropriate headings, following the structure of the original tool. For clarity and ease of reference, numerical identifiers were also assigned to each question, akin to the format used in the original checklist. Finally, the adapted questions were presented in a format that enables a straightforward evaluation, with response options of “yes,” “no,” or “not available”. The checklist included 19 questions pertaining to core MR assumptions, methods reporting, data presentation, interpretation and clinical implications (Adapted Critical Appraisal Tool—Supplementary Material).

The categorization of evidence robustness is based on a tool developed by Markozannes et al., who conducted a systematic review of MR studies on cancer risk. This categorization includes five predefined causality levels (robust, probable, suggestive, insufficient evidence or non-evaluable) [23]. It relies on data from both the primary MR analysis and at least one of the following methods: MR-Egger, weighted median (WM), MRPRESSO and Mendelian randomization with pleiotropy score. The analysis was done by two investigators (MA and BW). In studies 2 and 8, the results are considered robust when MR Egger is not included. Because it was found that MR Egger produced results that partially differed from other analytical methods, which can be attributed to differences in the underlying modelling techniques. In such cases, the robustness analyses both with and without MR Egger are provided.

This search yielded 530 potential studies. After removing duplicate entries, a total of 322 unique studies remained and 9 studies met the inclusion criteria and were thus analyzed in the systematic review. The search process and study selection are shown in a PRISMA flowchart (Fig. 1; [24]).

Out of the nine studies eight originated from Chinese research groups, with one coming from Sweden. The publications spanned from 2021 to 2023, with 6 studies conducted in the year 2023. The sample sizes (n) for the RA cases ranged from 12,838 to 462,933. The number of utilized single nucleotide polymorphisms (SNP) ranged from 8 to 142. All studies were conducted using European ancestry data and were using two-sample Mendelian randomization (TSMR). The analysis of the conducted sensitivity analyses showed that studies 4 and 5 demonstrated robust results, meeting the criteria of both statistical significance and directional consistency. Initially, studies 2 and 8 were categorized as probable but when MR Egger was not included, they also exhibited robustness; however, study 3 was rated as insufficient due to the absence of statistical significance in the findings. Despite variations in statistical significance across different methods, the remaining studies were classified as probable because they consistently showed directional consistency. Notably, none of the studies fell into the not evaluable category as all of them underwent a rigorous sensitivity analysis to ensure comprehensive assessment (Table 1).

In our adapted questionnaire, question 1 was only answered with “yes” in studies where outcomes were subsequently assessed as robust. This applies to studies 2, 4, 5, and 8. In study 4, the answer “no” was given for question 5, as this was interpreted to suggest the utilization of two distinct sample populations. This interpretation was based on the study’s description, which indicated the selection of genetic variants associated with RA from GWAS. Subsequently, genetic data for coronary atherosclerosis were obtained from the UK Biobank. In the questionnaire, the following question could not be sufficiently addressed in any of the studies: “Will interventions at a specific age yield effect of the same magnitude?” This inadequacy stemmed from either the absence of analysis or a lack of comprehensive discussion on this particular aspect within the studies. Question 9 was only answered with “no” in study 5 and question 14 was only answered with “no” in study 2 because the study did not adequately address the possibility of weak instrument bias or confounding through horizontal pleiotropy. In our opinion the study did not thoroughly explore or account for these potential sources of bias. These results are shown in Table 2. Comprehensive information, including the sizes of the case and SNP samples, specifications regarding the dataset employed as well as the crucial statistics such as odds ratios (OR), confidence intervals (CI), and p-values, can be readily accessed in Tables 3, 4 and 5. Tables 2, 3, 4 and 5 are providing a detailed overview of the essential data points, facilitating a more thorough understanding of the study’s findings and methodology. Table 3 includes information about the primary authors, the journal names and the publication years.

Overall, the included studies illustrate that genetically determined RA is causally associated with an increased risk of CVD, including myocardial infarction (MI), ischemic heart disease (IHD), and CAD. Furthermore, associations between the occurrence of RA and T2D (Study 9) and hypertension (Study 5) were also identified. The consistent results across various MR methods and sensitivity analyses support the credibility of these causal relationships. Specifically, it was found that RA is causally linked to CAD. Additionally, the studies emphasized the importance of considering systemic inflammation as a potential mechanistic link between RA and these diseases. On the other hand, no significant causal relationship between RA and certain cardiovascular outcomes, such as ischemic stroke, atrial fibrillation (AF), or arrhythmias, was observed (study 6). These results suggest that the genetic factors contributing to RA are not directly associated with rhythmological conditions. Regarding the association between heart failure and RA, there were varying results. Study 3 did not find sufficient evidence to support a causal relationship between genetically predicted RA and heart failure; however, study 7 indicated an association.

This systematic review examines the relationship between RA and CVD through the lens of MR studies. By analyzing nine selected MR studies, the study aimed to provide a comprehensive understanding of this complicated interaction, highlighting the genetic point of view. The findings consistently indicate a causal association between genetically determined RA and an increased risk of cardiovascular outcomes. This highlights the necessity of recognizing RA as a significant risk factor for CVD and underscores the importance of proactive cardiovascular risk management in RA patients. Furthermore, the review explores the genetic connections between RA and other cardiovascular risk factors such as type 2 diabetes (T2D) and hypertension, underscoring the systemic impact of RA beyond joint-related symptoms. A notable revelation from the reviewed studies is the potential mechanistic role of systemic inflammation, driven by genetic factors in mediating the RA-CVD relationship. Chronic inflammation is progressively seen as a substance in the development of atherosclerosis, suggesting that targeting inflammation might be a promising strategy to relieve cardiovascular risk in RA patients. The genetic mechanisms in RA should be more investigated in rheumatology to enhance understanding of this aspect of the disease’s development. Further research in this area is essential for advancing personalized treatment strategies and improving patient outcomes. The association between RA and CVD is likely more pronounced due to the interplay between inflammation and atherosclerosis. This interaction is not observed in the case of heart failure, suggesting different underlying mechanisms [25, 26].

For sure it is important to recognize variations in study results which highlight the need for further research to systematically clarify these genetic causal relationships. Potential mechanisms beyond chronic inflammation must be considered, including the role of the different RA treatments in modulating CVD risk.

Limitations of the reviewed studies include population specificity, potential pleiotropy and the need for validation in diverse populations, highlighting the importance of cautious interpretation and further investigation to confirm and expand upon the results. It must be mentioned that all genetic analyses were from European ancestries and that gender differences were not analyzed. The strengths of the study lie in its thorough examination of all MR studies from recent years and the study demonstrates a current and informed overview of the field. The analyzed studies have used an extensive dataset, which offers a substantial pool of data for analysis. Additionally, the study sheds light on a genetic connection that is currently not well understood. A risk of bias module was tailored specifically for the MR studies to analyze the risk of reverse causality offering robust evidence.

In summary, this systematic review of Mendelian randomization studies provides valuable insights into the complex interplay between rheumatoid arthritis and cardiovascular diseases focusing on the genetic aspects. The consistent findings across the selected MR studies reinforce the association between RA and elevated cardiovascular risk, particularly CAD. Despite limitations the studies underscore the necessity of proactive cardiovascular risk management in RA patients and the potential benefits of interventions targeting inflammation.