Bidirectional causal relationship between rheumatoid arthritis and systemic lupus erythematosus: a Mendelian randomization study

Background: Rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) are systemic, autoimmune inflammatory diseases. Although overlapping clinical and genetic features have been reported, the presence of a direct causal relationship between them remains uncertain. Methods: A bidirectional two-sample Mendelian randomization (MR) analysis was performed using genetic variants as instrumental variables to evaluate potential causal associations between RA and SLE. Sensitivity analyses were performed to assess heterogeneity, horizontal pleiotropy and the robustness of the findings. Results: Genetically predicted RA did not have a significant causal effect on susceptibility to SLE. Conversely, genetically predicted SLE did not demonstrate a significant causal effect on RA risk. Sensitivity analyses supported the stability and reliability of these findings, with no evidence of substantial heterogeneity or directional pleiotropy. Conclusion: We concluded that RA and SLE may share genetic susceptibility and environmental triggers without a unidirectional causal relationship.

Systemic lupus erythematosus, rheumatoid arthritis, Mendelian randomization, autoimmune diseases, genome-wide association study

Autoimmune diseases are characterized by immune system dysfunction, in which normal tissues and organs are mistakenly targeted and damaged [1,2]. A central feature of these disorders is the production of autoantibodies, which trigger sustained immune activation [3]. Rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) are prototypical autoimmune diseases that share overlapping immunopathogenic mechanisms and certain clinical manifestations [4]. The global incidence of SLE has been estimated at 43.7 (15.87-108.92)/100,000 individuals, with approximately 3.41 million affected worldwide [5]. This disease shows significant variation across sex, age, and some ethnic groups [6]. SLE is classically considered a type III hypersensitivity disorder and is characterized by the production of autoantibodies against double-stranded DNA, histones, nucleosomes and other chromatin components [7]. RA is a common systemic autoimmune disease of unclear etiology. It is characterized by the production of autoantibodies in the synovium, leading to chronic synovial inflammation and joint destruction [8]. In addition to synovial hyperplasia and persistent inflammation, RA may result in severe systemic complications [9]. Both SLE and RA are associated with systemic inflammation and multi-organ involvement, including the brain and vascular system [10]. Observational studies reported that some patients with RA may develop a lupus-like syndrome after penicillamine treatment [11,12]. Although RA and SLE exhibit overlapping clinical and serological features, observational studies cannot determine whether their co-occurrence reflects shared genetic and environmental susceptibility or a direct causal influence of one disease on the other. In this study, this limitation was addressed by applying Mendelian randomization (MR) to evaluate the potential lifelong causal relationship between RA and SLE. MR is based on the random allocation of single nucleotide polymorphism (SNP) at conception [13]. Genetic variants are randomly inherited from parents and are generally independent of postnatal environmental influences [14]. This approach relies on three core assumptions: genetic variants are associated with the exposure of interest and are not associated with confounders, and they influence the outcome only through exposure [15,16]. By leveraging genetic associations, potential causal relationships between exposures and outcomes can be inferred [17]. MR reduces confounding factors and reverse causation compared with traditional observational studies, thereby strengthening causal inference [18]. Additionally, publicly available genome-wide association data can be used for two-sample MR analyses [19]. Accordingly, a bidirectional two-sample MR analysis was performed to investigate the potential causal relationship between SLE and RA.

Patients Thirteen patients with SLE, thirteen patients with RA, and thirteen healthy volunteers aged between 18 and 84 years were recruited at the Affiliated Hospital of Nantong University for this investigation. The study received approval from the Ethics Committee of the Affiliated Hospital of Nantong University (2017-K003). Fresh peripheral blood and serum samples, each totaling 5 mL, were collected from individuals diagnosed with SLE or RA, as well as from healthy subjects, for the purpose of conducting quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) and enzyme-linked immunosorbent assay (ELISA). qRT-PCR assay Total RNA was isolated utilizing TRIzol Reagent (15596026, Invitrogen). Approximately 500 ng of RNA was obtained from each sample for the reverse transcription process. The PrimeScript™ II 1st Strand cDNA Synthesis Kit (6210A, Takara) was employed for reverse transcription, resulting in the synthesis of the first strand cDNA. TB Green Premix Ex TaqII (RR820, Takara) served as the fluorescent reporter dye, and qPCR analysis was conducted using the LightCycler® 96 Real-Time PCR System (Roche). Primer sequences were as follows: Signal transducer and activator of transcription 4 (STAT4)-F: AAGGGATGGGTAGCCAGGAT; STAT4-R: AGGAGGCTAGGTCAGGAAGG; Interferon regulatory factor 5 (IRF5)-F: TCTTCCTCCTCCTCTCCTGC; IRF5-R: GGACTTCCGCCTCATCTACG; Glyceraldehyde-3-phosphate dehydrogenase (GAPDH)-F: GTCAAGGCTGAGAACGGGAA; GAPDH-R: AAATGAGCCCCAGCCTTCTC. Tumor necrosis factor-α (TNF-α) assay The serum levels of TNF-α in patients with SLE or RA, along with those in healthy volunteers, were assessed utilizing a commercial ELISA kit (MULTI SCIENCES, EK182) in accordance with the manufacturer’s guidelines. Source of GWAS data for exposure and outcome Genome-wide association study (GWAS) datasets derived from comparable populations were selected for both exposure and outcome analyses to ensure balanced sex representation and minimize population stratification bias. The GWAS dataset for SLE (ebi-a-GCST003156) was obtained from the IEU OpenGWAS database, which included 14,267 samples and 7,071,163 SNPs (https://GWAS.mrcieu.ac.uk/). The GWAS dataset for RA (ebi-a-GCST005569) was retrieved from the same database, with 47,580 samples and 112,654 SNPs. Detailed information on the selected GWAS datasets is summarized in Table 1. Table 1. | Systemic Lupus Erythematosus | Rheumatoid Arthritis | | |---|---|---| | Population | European | European | | Ncase | 5201 | 13838 | | Ncontrol | 9066 | 33742 | | Sample size | 14267 | 47580 | | SNPs | 7071163 | 112654 | | GWAS ID | ebi-a-GCST003156 | ebi-a-GCST005569 | GWAS, genome-wide association studies; Ncase, number of cases; Ncontrol, number of controls; SNPs, single-nucleotide polymorphisms; SLE, systemic lupus erythematosus; RA, rheumatoid arthritis. Bidirectional MR analysis A bidirectional univariate MR analysis was performed to evaluate the causal association between SLE and RA. SNPs strongly associated with the exposure were selected using a genome-wide significance threshold (P < 5 × 10-8). SNPs were pruned using independence criteria (r2 < 0.001, distance = 10,000 kb) to minimize bias due to linkage disequilibrium (LD). Selected SNPs were further screened in PhenoScanner (www.PhenoScanner.medschl.cam.ac.uk) and SNPs associated with potential confounding factors or outcomes were excluded. Instrument strength was assessed using the F-statistic (F = R2(N-1-k)/(1-R2)k or F = β2/se2) to reduce weak instrument bias [20,21]. SNP-outcome effect estimates were extracted from the corresponding GWAS dataset, with a minor allele frequency (MAF) threshold > 0.01. Causal estimates were calculated using multiple MR methods, including multiplicative random effects-inverse variance weighted (MRE-IVW), inverse variance weighted (IVW), weighted median (WM), and MR Egger (ME) regression. The IVW method was considered the primary analysis under the assumptions that all instrumental variables satisfied the core MR assumptions. The WM method provides reliable estimates when at least 50% of the weight originates from valid instruments [22]. ME allows the detection of directional pleiotropy through its intercept term, providing adjusted estimates when pleiotropy is present. The RE-IVW method accounts for heterogeneity among instruments, yielding more conservative estimates when heterogeneity is present. Statistical analysis Heterogeneity among instrumental variables was assessed using Cochran Q statistic and visualized with funnel plots [23,24]. The MRE-IVW model was applied when significant heterogeneity was detected (P < 0.05). A leave-one-out analysis was performed to evaluate the influence of individual SNPs on the overall estimates and to assess result stability [25]. Horizontal pleiotropy was examined using the ME intercept test. A statistically significant intercept (P < 0.05) indicated potential directional pleiotropy [26]. A flow chart of the study design is represented in Figure 1. Functional experimental data are depicted as mean + standard error of the mean (SEM). Statistical analysis was carried out using GraphPad Prism software Version 6.01 (GraphPad Inc., La Jolla, CA, USA). One-way ANOVA was utilized for comparisons among multiple groups with Dunnett’s comparisons test. Significance levels were denoted as follows: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Extraction of instrumental variables for MR A bidirectional two-sample MR analysis was performed. Independent SNPs strongly associated with the exposure were identified within a 10,000 kb window using a genome-wide significant threshold of P < 5 × 10-8, prioritizing SNPs showing low correlation with other SNPs in the region (LD, r2 < 0.001). These SNPs were screened in PhenoScanner to remove those associated with potential confounders and outcomes. Common risk factors for SLE included, but were not limited to smoking, alcohol consumption, obesity, endometriosis, silica exposure, air pollution, infections, and vitamin D deficiency [27,28]. Established risk factors for RA included smoking, unhealthy lifestyle choices, dietary factors, and infections [29]. When RA was treated as the exposure variable, 15 SNPs significantly associated with RA were initially identified. After PhenoScanner screening, rs11758312, rs13426947, and rs6679677 were excluded due to strong associations with the outcome. Additionally, rs3129886, rs660895, rs6920220, and rs71624119 were excluded because they were associated with immune diseases, including Crohn’s disease and inflammatory bowel disease. Ultimately, 8 instrumental variables were identified for MR analysis (Table 2). In the reverse MR analysis, SLE was considered as the exposure variable, and two SNPs, namely rs58721818 and rs6679677, were removed using the same exclusion criteria previously described (Table 3), and 13 instrumental variables were retained for further analysis. Table 2. | SNP | Expose | Outcome | Chr | EA | OA | Β (expose) | Β (outcome) | SE (expose) | SE (outcome) | EAF | P (expose) | P (outcome) | KEEP | F-statistic | |---|---|---|---|---|---|---|---|---|---|---|---|---|---|---| | rs10209110 | RA | SLE | 2 | T | C | -0.1058 | -0.04085 | 0.018527 | 0.029876 | NA | 1.13E-08 | 0.171476 | TRUE | 32.61042 | | rs17630466 | RA | SLE | 4 | G | A | 0.126721 | -0.0202 | 0.020018 | 0.02481 | NA | 2.45E-10 | 0.415467 | TRUE | 40.07417 | | rs2228145 | RA | SLE | 1 | C | A | -0.1087 | 0.04879 | 0.019124 | 0.028077 | NA | 1.32E-08 | 0.082262 | TRUE | 32.30745 | | rs2240339 | RA | SLE | 1 | T | C | -0.1125 | 0.00995 | 0.01981 | 0.028124 | NA | 1.36E-08 | 0.723486 | TRUE | 32.24778 | | rs2812378 | RA | SLE | 9 | A | G | -0.11038 | 0.02020 | 0.019415 | 0.030501 | NA | 1.31E-08 | 0.507745 | TRUE | 32.32107 | | rs3129294 | RA | SLE | 6 | C | A | -0.17698 | -0.04082 | 0.01994 | 0.033437 | NA | 6.96E-19 | 0.222144 | TRUE | 78.77635 | | rs8026898 | RA | SLE | 15 | A | G | 0.124604 | -0.01005 | 0.020834 | 0.029432 | NA | 2.22E-09 | 0.732745 | TRUE | 35.7686 | | rs9262218 | RA | SLE | 6 | T | C | -0.44723 | 0.10436 | 0.069481 | 0.086959 | NA | 1.22E-10 | 0.230098 | TRUE | 41.43108 | SNPs, single-nucleotide polymorphisms; Chr, chromosome; EA, effect allele; OA, other allele; SE, standard error; EAF, effect allele frequency; β, beta coefficient; SLE, systemic lupus erythematosus; RA, rheumatoid arthritis; P, P-value; F-statistic, Fisher’s F-statistic. Table 3. | SNP | Expose | Outcome | Chr | EA | OA | Β (expose) | Β (outcome) | SE (expose) | SE (outcome) | EAF | P (expose) | P (outcome) | KEEP | F-statistic | |---|---|---|---|---|---|---|---|---|---|---|---|---|---|---| | rs1143679 | SLE | RA | 16 | A | G | 0.58221 | 0.038451 | 0.039987 | 0.029505 | NA | 5.03E-48 | 0.1925 | TRUE | 212.0017 | | rs17849501 | SLE | RA | 1 | T | C | 0.81093 | 0.071483 | 0.049864 | 0.062535 | NA | 1.81E-59 | 0.253 | TRUE | 264.4777 | | rs2431697 | SLE | RA | 5 | C | T | -0.22314 | -0.03376 | 0.029296 | 0.018771 | NA | 2.60E-14 | 0.072059 | TRUE | 58.01522 | | rs2459611 | SLE | RA | 2 | T | C | 0.26136 | 0.036249 | 0.045245 | 0.030838 | NA | 7.62E-09 | 0.2398 | TRUE | 33.36981 | | rs35000415 | SLE | RA | 7 | T | C | 0.58778 | 0.11876 | 0.041539 | 0.0279 | NA | 1.86E-45 | 2.08E-05 | TRUE | 200.2294 | | rs353608 | SLE | RA | 11 | G | A | 0.18633 | -0.02946 | 0.02802 | 0.018569 | NA | 2.93E-11 | 0.1126 | TRUE | 44.22171 | | rs389884 | SLE | RA | 6 | G | A | 0.92821 | -0.06507 | 0.043232 | 0.0291 | NA | 2.92E-10 | 0.02534 | TRUE | 460.9908 | | rs4916215 | SLE | RA | 1 | T | C | 0.22314 | 0.053612 | 0.033969 | 0.02116 | NA | 5.07E-11 | 0.01129 | TRUE | 43.15163 | | rs58688157 | SLE | RA | 11 | G | A | -0.22314 | -0.01177 | 0.033565 | 0.020995 | NA | 2.97E-11 | 0.5751 | TRUE | 44.19823 | | rs6889239 | SLE | RA | 5 | C | T | 0.27763 | 0.070086 | 0.03174 | 0.021041 | NA | 2.19E-18 | 0.000866 | TRUE | 76.51123 | | rs7097397 | SLE | RA | 10 | A | G | -0.18633 | 0.006876 | 0.028712 | 0.019429 | NA | 8.60E-11 | 0.723401 | TRUE | 42.11577 | | rs73068668 | SLE | RA | 19 | A | G | -0.31471 | -0.01765 | 0.05749 | 0.035769 | NA | 4.40E-08 | 0.6216 | TRUE | 29.96641 | | rs7768653 | SLE | RA | 6 | T | C | -0.20701 | -0.04421 | 0.029689 | 0.018744 | NA | 3.11E-12 | 0.01835 | TRUE | 48.61893 | SNPs, single-nucleotide polymorphisms; Chr, chromosome; EA, effect allele; OA, other allele; SE, standard error; EAF, effect allele frequency; β, beta coefficient; SLE, systemic lupus erythematosus; RA, rheumatoid arthritis; P, P-value; F-statistic, Fisher’s F-statistic. SNP effect estimates were extracted from the corresponding outcome GWAS datasets and filtered using MAF threshold > 0.01. After harmonization of exposure and outcome alleles, as well as exclusion of weak instrument F-statistics < 10 or incongruent SNPs, effective instrumental variables were identified and included in MR analysis using MRE-IVW, IVW, WM and ME regression methods. Causal effects of RA on SLE Eight instrumental variables were identified to evaluate the causal association between RA (exposure) and SLE (outcome), and the P values from the MR analyses exceeded 0.05 (IVW: OR = 0.931, 95% CI 0.78-1.11, P = 0.419; ME: OR = 0.899, 95% CI 0.52-1.57, P = 0.720; WM: OR = 0.884, 95% CI 0.72-1.09, P = 0.245; MRE-IVW: OR = 0.931, 95% CI 0.78-1.11, P = 0.419), indicating a lack of significant correlation (Table 4 and Figure 2). Table 4. | Exposure | Outcome | Method | nSNP | OR | 95% CI | P value | |---|---|---|---|---|---|---| | RA | SLE | IVW | 8 | 0.931 | 0.78-1.11 | 0.419 | | ME | 0.899 | 0.52-1.57 | 0.720 | ||| | WM | 0.884 | 0.72-1.09 | 0.245 | ||| | MRE-IVW | 0.931 | 0.78-1.11 | 0.419 | ||| | SLE | RA | IVW | 13 | 1.061 | 0.99-1.14 | 0.100 | | ME | 0.992 | 0.87-1.13 | 0.910 | ||| | WM | 1.063 | 0.99-1.14 | 0.105 | ||| | MRE-IVW | 1.061 | 0.99-1.14 | 0.100 | RA, rheumatoid arthritis; SLE, systemic lupus erythematosus; SNPs, single-nucleotide polymorphisms; IVW, inverse variance weighted; ME, MR-Egger; WM, weighted median; MRE-IVW, multiplicative random effects-inverse variance weighted; OR, odds ratio; CI, confidence interval; P > 0.05. Causal effects of SLE on RA A total of 13 SNPs were chosen for evaluation through the reverse MR analysis, with the outcomes (Table 4 and Figure 2) also yielding non-significant results (IVW: OR = 1.061, 95% CI 0.99-1.14, P = 0.100; ME: OR = 0.992, 95% CI 0.87-1.13, P = 0.910; WM: OR = 1.063, 95% CI 0.99-1.14, P = 0.105; MRE-IVW: OR = 1.061, 95% CI 0.99-1.14, P = 0.100). Sensitivity analysis F-statistics were calculated for each instrumental variable to assess instrument strength, and all selected variants exceeded the threshold of 10 (Tables 2 and 3). Heterogeneity among instrumental variables used in the present MR analysis was evaluated using Cochran’s Q statistics under both MR-Egger (ME) and inverse variance weighted (IVW) models. No significant heterogeneity was found for the effect of RA on SLE (ME: value = 0.21; IVW: Q-P value = 0.30) (Table 5). However, significant heterogeneity was observed for the effect of SLE on RA (ME: Q-P value = 2.30 × 10-5; IVW: Q-P value = 5.59 × 10-6) (Table 5). Funnel plots demonstrated a symmetrical distribution of instrumental variables for the impact of RA on SLE, but not for the impact of SLE on RA (Figure 3), consistent with the heterogeneity findings. The MRE-IVW method was prioritized for the SLE on RA analysis due to the presence of significant heterogeneity coupled with funnel plot asymmetry, as this method accounts for between-instrument variability and provides more conservative estimates. The MRE-IVW analysis yielded an OR of 1.061 (95% CI 0.99-1.14; P = 0.100) indicating no statistically significant causal effect of SLE on RA. This finding was the most reliable estimate. Table 5. | Exposure | Outcome | Method | Cochran’s Q | Q-df | Q-P-value | |---|---|---|---|---|---| | RA | SLE | ME | 8.36 | 6 | 0.21 | | IVW | 8.38 | 7 | 0.30 | || | SLE | RA | ME | 41.12 | 11 | 2.30 × 10-5 | | IVW | 46.54 | 12 | 5.59 × 10-6 | RA, rheumatoid arthritis; SLE, systemic lupus erythematosus; ME, MR-Egger; IVW, Inverse variance weighted; Q-df, Q-degrees of freedom; Q-P value, P value of Cochran’s Q test; Q-P value (RA on SLE) > 0.05; Q-P value (SLE on RA) < 0.05. Pleiotropy was assessed using the ME intercept test, which indicated no evidence of directional pleiotropy in the instrumental variables employed to explore the causal relationship between RA and SLE (Table 6). Additionally, a leave-one-out analysis was performed to evaluate the influence of individual SNPs on the overall causal estimation by systematically excluding each SNP and reiterating the MR analyses. The results remained consistent even after sequential removal of each SNP, as illustrated in Figure 4. Table 6. | Exposure | Outcome | MR-Egger-intercept | SE | P value | |---|---|---|---|---| | RA | SLE | 5.12E-03 | 0.038 | 0.898 | | SLE | RA | 0.028 | 0.023 | 0.254 | RA, rheumatoid arthritis; SLE, systemic lupus erythematosus; SE, standard error; MR, Mendelian randomization; P > 0.05. Relevant indicator evaluation in RA and SLE We analyzed serum TNF-α, a pivotal inflammatory cytokine implicated in the pathophysiology of both RA and SLE [30,31]. As anticipated, patients with RA or SLE exhibited markedly elevated circulating TNF-α compared with healthy controls (Figure 5A), reaffirming the immune-mediated inflammatory nature of both diseases. We next investigated two biomarkers - IRF5 and STAT4 - that have previously been linked to SLE and RA [32-34]. qRT-PCR analysis on peripheral blood showed that both IRF5 and STAT4 mRNA expression levels were significantly up-regulated in individuals with SLE and RA relative to healthy controls (Figure 5B and 5C). Integrating these findings with our MR analysis, the convergence of elevated TNF-α, IRF5, and STAT4 points to a common inflammatory network underlying RA and SLE, despite the absence of a direct causal relationship between the two conditions at the genetic level.

This study evaluated the potential bidirectional direct causal relationship between RA and SLE using MR framework. No significant evidence of a direct genetic causal association was identified in either direction. These findings suggested that the observed clinical overlap between RA and SLE was unlikely to be explained by a unidirectional genetic causal pathway. RA and SLE share several clinical manifestations, including joint pain, swelling, systemic inflammation, and multi-organ involvement [35,36]. Sequential or overlapping presentations have been reported in some patients, and the coexistence of both conditions is termed ‘Rhupus’ [37,38]. Furthermore, overlapping genetic susceptibility loci have been described, indicating a potential connection between the two diseases [39,40]. However, pathogenic and biological mechanisms are different. SLE is characterized by dysregulated autoantibody production and complement activity, while RA is driven largely by synovial inflammation mediated by cytokines and macrophages [41,42]. These distinct mechanisms may explain why no direct genetic causal relationship was detected. Several interpretations may be considered. First, shared genetic susceptibility loci may increase overall autoimmune predisposition without establishing a direct causal pathway between the two diseases [43]. Many of these loci encode broadly acting immunomodulatory molecules rather than disease-specific effectors. These gene mutations may simultaneously increase the body’s susceptibility to multiple autoimmune reactions, but do not establish a unidirectional and inevitable causal relationship between RA and SLE. Second, autoimmune diseases arise from complex interactions between genetic background and environmental exposure [3]. Environmental triggers may differentially influence disease expression in individuals with similar genetic backgrounds [44]. For example, individuals carrying specific HLA genotypes are associated with RA in the context of smoking [45], while ultraviolet exposure has been linked to SLE under the same genotype [46]. Such gene-environment interaction may lead to temporal overlap in the same individual, without implying genetic causation. Epigenetic regulation may further mediate the interaction between environmental exposure and immune activation [47]. Persistent immune dysregulation may facilitate autoimmune responses against different self-antigens, potentially contributing to overlapping phenotypes such as Rhupus syndrome. These observations support the hypothesis that Rhupus represents a distinct immune endotype rather than a linear progression from one disease to another. Such patients may carry a specific genetic background (such as a specific HLA haplotype combination) that predisposes their immune system to produce a broad autoimmune response attacking both joints and multiple organs [48]. Thus, the interaction between multiple genes and various factors likely contributes to the development of both diseases [49]. From a clinical perspective, the absence of a direct genetic causal relationship suggests that RA does not inherently confer genetic risk for SLE and vice versa. Thus, Rhupus syndrome is a unique disease endotype, rather than a simple superposition of two diseases. Therefore, routine management should remain disease-specific, while recognizing that shared susceptibility and environmental factors may contribute to overlapping clinical features. Taken together, this study provided a theoretical basis for more precise risk assessment, diagnostic stratification, and personalized treatment. The application of bidirectional two-sample MR analysis strengthens causal inference by reducing confounding and reverse causation [50]. Besides, MR was not used to analyze individual samples but the whole population, supporting the statistical power of the results. Several limitations of this research should be acknowledged. The analysis was primarily based on individuals of European ancestry, which may limit generalizability to other populations with different genetic architectures, population-specific allele frequencies, and linkage disequilibrium patterns, and disease prevalence may influence causal estimates. For example, the frequency of the risk allele T of STAT4 rs7574865 is higher in Asian populations than in European populations [51]. The prevalence of SLE is not only higher in Asian populations, but the condition is often more severe [52]. Moreover, there are differences in the linkage disequilibrium structure. The instrumental variable SNP that is effective in the European population may be invalid in other populations due to changes in the strength of association with the pathogenic site [53], thus weakening the accuracy of the analysis. In addition, gene-environment interactions are not fully captured by standard MR approaches. Finally, the size of GWAS datasets used in this study was small, and causal effect estimates in MR analysis usually increase with larger sample sizes. Larger GWAS datasets and more diverse ancestry representation would improve statistical power and external validity. In conclusion, no direct bidirectional genetic causal relationship between RA and SLE was identified. The clinical overlap between the two diseases is more likely attributable to shared genetic susceptibility, environmental influences, and complex immune regulatory mechanisms rather than a simple unidirectional causal pathway. Further study of the interaction between several genes and environmental factors may be helpful in discussing the potential correlation between RA and SLE.

This work was supported in part by Jiangsu Provincial Research Hospital (YJXYY202204). Disclosure of conflict of interest None.

- 1.Xiao ZX, Miller JS, Zheng SG. An updated advance of autoantibodies in autoimmune diseases. Autoimmun Rev. 2021;20:102743. doi: 10.1016/j.autrev.2020.102743. [DOI] [PubMed] [Google Scholar] - 2.Yang S, Zhao M, Jia S. Macrophage: key player in the pathogenesis of autoimmune diseases. Front Immunol. 2023;14:1080310. doi: 10.3389/fimmu.2023.1080310. [DOI] [PMC free article] [PubMed] [Google Scholar] - 3.Pisetsky DS. Pathogenesis of autoimmune disease. Nat Rev Nephrol. 2023;19:509–524. doi: 10.1038/s41581-023-00720-1. [DOI] [PMC free article] [PubMed] [Google Scholar] - 4.Restivo V, Candiloro S, Daidone M, Norrito R, Cataldi M, Minutolo G, Caracci F, Fasano S, Ciccia F, Casuccio A, Tuttolomondo A. Systematic review and meta-analysis of cardiovascular risk in rheumatological disease: symptomatic and non-symptomatic events in rheumatoid arthritis and systemic lupus erythematosus. Autoimmun Rev. 2022;21:102925. doi: 10.1016/j.autrev.2021.102925. [DOI] [PubMed] [Google Scholar] - 5.Tian J, Zhang D, Yao X, Huang Y, Lu Q. Global epidemiology of systemic lupus erythematosus: a comprehensive systematic analysis and modelling study. Ann Rheum Dis. 2023;82:351–356. doi: 10.1136/ard-2022-223035. [DOI] [PMC free article] [PubMed] [Google Scholar] - 6.Allen ME, Rus V, Szeto GL. Leveraging heterogeneity in systemic lupus erythematosus for new therapies. Trends Mol Med. 2021;27:152–171. doi: 10.1016/j.molmed.2020.09.009. [DOI] [PMC free article] [PubMed] [Google Scholar] - 7.Rekvig OP, Van der Vlag J. The pathogenesis and diagnosis of systemic lupus erythematosus: still not resolved. Semin Immunopathol. 2014;36:301–311. doi: 10.1007/s00281-014-0428-6. [DOI] [PubMed] [Google Scholar] - 8.Weyand CM, Goronzy JJ. The immunology of rheumatoid arthritis. Nat Immunol. 2021;22:10–18. doi: 10.1038/s41590-020-00816-x. [DOI] [PMC free article] [PubMed] [Google Scholar] - 9.McInnes IB, Schett G. The pathogenesis of rheumatoid arthritis. N Engl J Med. 2011;365:2205–2219. doi: 10.1056/NEJMra1004965. [DOI] [PubMed] [Google Scholar] - 10.Zhernakova A, Withoff S, Wijmenga C. Clinical implications of shared genetics and pathogenesis in autoimmune diseases. Nat Rev Endocrinol. 2013;9:646–659. doi: 10.1038/nrendo.2013.161. [DOI] [PubMed] [Google Scholar] - 11.Chalmers A, Thompson D, Stein HE, Reid G, Patterson AC. Systemic lupus erythematosus during penicillamine therapy for rheumatoid arthritis. Ann Intern Med. 1982;97:659–663. doi: 10.7326/0003-4819-97-5-659. [DOI] [PubMed] [Google Scholar] - 12.Borg AA, Davis MJ, Dawes PT, Shadforth MF. Combination therapy for rheumatoid arthritis and drug-induced systemic lupus erythematosus. Clin Rheumatol. 1994;13:522–524. doi: 10.1007/BF02242956. [DOI] [PubMed] [Google Scholar] - 13.Taylor AE, Davies NM, Ware JJ, VanderWeele T, Smith GD, Munafò MR. Mendelian randomization in health research: using appropriate genetic variants and avoiding biased estimates. Econ Hum Biol. 2014;13:99–106. doi: 10.1016/j.ehb.2013.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar] - 14.Lawlor DA, Harbord RM, Sterne JAC, Timpson N, Davey Smith G. Mendelian randomization: using genes as instruments for making causal inferences in epidemiology. Stat Med. 2008;27:1133–1163. doi: 10.1002/sim.3034. [DOI] [PubMed] [Google Scholar] - 15.Castel SE, Cervera A, Mohammadi P, Aguet F, Reverter F, Wolman A, Guigo R, Iossifov I, Vasileva A, Lappalainen T. Modified penetrance of coding variants by cis-regulatory variation contributes to disease risk. Nat Genet. 2018;50:1327–1334. doi: 10.1038/s41588-018-0192-y. [DOI] [PMC free article] [PubMed] [Google Scholar] - 16.Bennett DA. An introduction to instrumental variables--part 2: mendelian randomisation. Neuroepidemiology. 2010;35:307–310. doi: 10.1159/000321179. [DOI] [PubMed] [Google Scholar] - 17.Verduijn M, Siegerink B, Jager KJ, Zoccali C, Dekker FW. Mendelian randomization: use of genetics to enable causal inference in observational studies. Nephrol Dial Transplant. 2010;25:1394–1398. doi: 10.1093/ndt/gfq098. [DOI] [PubMed] [Google Scholar] - 18.Evans DM, Davey Smith G. Mendelian randomization: new applications in the coming age of hypothesis-free causality. Annu Rev Genomics Hum Genet. 2015;16:327–350. doi: 10.1146/annurev-genom-090314-050016. [DOI] [PubMed] [Google Scholar] - 19.Pierce BL, Burgess S. Efficient design for Mendelian randomization studies: subsample and 2-sample instrumental variable estimators. Am J Epidemiol. 2013;178:1177–1184. doi: 10.1093/aje/kwt084. [DOI] [PMC free article] [PubMed] [Google Scholar] - 20.Xu Q, Ni JJ, Han BX, Yan SS, Wei XT, Feng GJ, Zhang H, Zhang L, Li B, Pei YF. Causal relationship between gut microbiota and autoimmune diseases: a two-sample Mendelian randomization study. Front Immunol. 2022;12:746998. doi: 10.3389/fimmu.2021.746998. [DOI] [PMC free article] [PubMed] [Google Scholar] - 21.Dan YL, Wang P, Cheng Z, Wu Q, Wang XR, Wang DG, Pan HF. Circulating adiponectin levels and systemic lupus erythematosus: a two-sample Mendelian randomization study. Rheumatology (Oxford) 2021;60:940–946. doi: 10.1093/rheumatology/keaa506. [DOI] [PubMed] [Google Scholar] - 22.Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol. 2017;32:377–389. doi: 10.1007/s10654-017-0255-x. [DOI] [PMC free article] [PubMed] [Google Scholar] - 23.Huedo-Medina TB, Sánchez-Meca J, Marín-Martínez F, Botella J. Assessing heterogeneity in meta-analysis: Q statistic or I2 index? Psychol Methods. 2006;11:193–206. doi: 10.1037/1082-989X.11.2.193. [DOI] [PubMed] [Google Scholar] - 24.Gill D. Heterogeneity between genetic variants as a proxy for pleiotropy in Mendelian randomization. JAMA Cardiol. 2020;5:107–108. doi: 10.1001/jamacardio.2019.4281. [DOI] [PubMed] [Google Scholar] - 25.Mokry LE, Ross S, Timpson NJ, Sawcer S, Davey Smith G, Richards JB. Obesity and multiple sclerosis: a Mendelian randomization study. PLoS Med. 2016;13:e1002053. doi: 10.1371/journal.pmed.1002053. [DOI] [PMC free article] [PubMed] [Google Scholar] - 26.Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50:693–698. doi: 10.1038/s41588-018-0099-7. [DOI] [PMC free article] [PubMed] [Google Scholar] - 27.Taylor PN, Albrecht D, Scholz A, Gutierrez-Buey G, Lazarus JH, Dayan CM, Okosieme OE. Global epidemiology of hyperthyroidism and hypothyroidism. Nat Rev Endocrinol. 2018;14:301–316. doi: 10.1038/nrendo.2018.18. [DOI] [PubMed] [Google Scholar] - 28.Gergianaki I, Bortoluzzi A, Bertsias G. Update on the epidemiology, risk factors, and disease outcomes of systemic lupus erythematosus. Best Pract Res Clin Rheumatol. 2018;32:188–205. doi: 10.1016/j.berh.2018.09.004. [DOI] [PubMed] [Google Scholar] - 29.Venetsanopoulou AI, Alamanos Y, Voulgari PV, Drosos AA. Epidemiology of rheumatoid arthritis: genetic and environmental influences. Expert Rev Clin Immunol. 2022;18:923–931. doi: 10.1080/1744666X.2022.2106970. [DOI] [PubMed] [Google Scholar] - 30.Miyazaki Y, Nakano K, Nakayamada S, Kubo S, Iwata S, Hanami K, Fukuyo S, Miyagawa I, Yamaguchi A, Kawabe A, Saito K, Tanaka Y. Serum TNFα levels at 24 h after certolizumab pegol predict effectiveness at week 12 in patients with rheumatoid arthritis from TSUBAME study. Arthritis Res Ther. 2021;23:154. doi: 10.1186/s13075-021-02547-2. [DOI] [PMC free article] [PubMed] [Google Scholar] - 31.Weckerle CE, Mangale D, Franek BS, Kelly JA, Kumabe M, James JA, Moser KL, Harley JB, Niewold TB. Large-scale analysis of tumor necrosis factor α levels in systemic lupus erythematosus. Arthritis Rheum. 2012;64:2947–2952. doi: 10.1002/art.34483. [DOI] [PMC free article] [PubMed] [Google Scholar] - 32.Abelson AK, Delgado-Vega AM, Kozyrev SV, Sánchez E, Velázquez-Cruz R, Eriksson N, Wojcik J, Linga Reddy MV, Lima G, D’Alfonso S, Migliaresi S, Baca V, Orozco L, Witte T, Ortego-Centeno N, Abderrahim H, Pons-Estel BA, Gutiérrez C, Suárez A, González-Escribano MF, Martin J, Alarcón-Riquelme ME. STAT4 associates with systemic lupus erythematosus through two independent effects that correlate with gene expression and act additively with IRF5 to increase risk. Ann Rheum Dis. 2009;68:1746–1753. doi: 10.1136/ard.2008.097642. [DOI] [PMC free article] [PubMed] [Google Scholar] - 33.Seyhan AA, Gregory B, Cribbs AP, Bhalara S, Li Y, Loreth C, Zhang Y, Guo Y, Lin LL, Feldmann M, Williams LM, Brennan FM, Taylor PC. Novel biomarkers of a peripheral blood interferon signature associated with drug-naïve early arthritis patients distinguish persistent from self-limiting disease course. Sci Rep. 2020;10:8830. doi: 10.1038/s41598-020-63757-3. [DOI] [PMC free article] [PubMed] [Google Scholar] - 34.Frucht DM, Aringer M, Galon J, Danning C, Brown M, Fan S, Centola M, Wu CY, Yamada N, El Gabalawy H, O’Shea JJ. Stat4 is expressed in activated peripheral blood monocytes, dendritic cells, and macrophages at sites of Th1-mediated inflammation. J Immunol. 2000;164:4659–4664. doi: 10.4049/jimmunol.164.9.4659. [DOI] [PubMed] [Google Scholar] - 35.Tani C, D’Aniello D, Delle Sedie A, Carli L, Cagnoni M, Possemato N, Carbone M, Della Rossa A, Riente L, Baldini C, Talarico R, Caramella D, Bombardieri S, Mosca M. Rhupus syndrome: assessment of its prevalence and its clinical and instrumental characteristics in a prospective cohort of 103 SLE patients. Autoimmun Rev. 2013;12:537–541. doi: 10.1016/j.autrev.2012.09.004. [DOI] [PubMed] [Google Scholar] - 36.Chi XK, Xu XL, Chen BY, Su J, Du YZ. Combining nanotechnology with monoclonal antibody drugs for rheumatoid arthritis treatments. J Nanobiotechnology. 2023;21:105. doi: 10.1186/s12951-023-01857-8. [DOI] [PMC free article] [PubMed] [Google Scholar] - 37.Cohen MG, Webb J. Concurrence of rheumatoid arthritis and systemic lupus erythematosus: report of 11 cases. Ann Rheum Dis. 1987;46:853–858. doi: 10.1136/ard.46.11.853. [DOI] [PMC free article] [PubMed] [Google Scholar] - 38.Antonini L, Le Mauff B, Marcelli C, Aouba A, de Boysson H. Rhupus: a systematic literature review. Autoimmun Rev. 2020;19:102612. doi: 10.1016/j.autrev.2020.102612. [DOI] [PubMed] [Google Scholar] - 39.Remmers EF, Plenge RM, Lee AT, Graham RR, Hom G, Behrens TW, de Bakker PIW, Le JM, Lee HS, Batliwalla F, Li W, Masters SL, Booty MG, Carulli JP, Padyukov L, Alfredsson L, Klareskog L, Chen WV, Amos CI, Criswell LA, Seldin MF, Kastner DL, Gregersen PK. STAT4 and the risk of rheumatoid arthritis and systemic lupus erythematosus. N Engl J Med. 2007;357:977–986. doi: 10.1056/NEJMoa073003. [DOI] [PMC free article] [PubMed] [Google Scholar] - 40.Márquez A, Vidal-Bralo L, Rodríguez-Rodríguez L, González-Gay MA, Balsa A, González-Álvaro I, Carreira P, Ortego-Centeno N, Ayala-Gutiérrez MM, García-Hernández FJ, González-Escribano MF, Sabio JM, Tolosa C, Suárez A, González A, Padyukov L, Worthington J, Vyse T, Alarcón-Riquelme ME, Martín J. A combined large-scale meta-analysis identifies COG6 as a novel shared risk locus for rheumatoid arthritis and systemic lupus erythematosus. Ann Rheum Dis. 2017;76:286–294. doi: 10.1136/annrheumdis-2016-209436. [DOI] [PubMed] [Google Scholar] - 41.Marion TN, Postlethwaite AE. Chance, genetics, and the heterogeneity of disease and pathogenesis in systemic lupus erythematosus. Semin Immunopathol. 2014;36:495–517. doi: 10.1007/s00281-014-0440-x. [DOI] [PubMed] [Google Scholar] - 42.McGonagle D, Watad A, Savic S. Mechanistic immunological based classification of rheumatoid arthritis. Autoimmun Rev. 2018;17:1115–1123. doi: 10.1016/j.autrev.2018.06.001. [DOI] [PubMed] [Google Scholar] - 43.Orozco G, Eyre S, Hinks A, Bowes J, Morgan AW, Wilson AG, Wordsworth P, Steer S, Hocking L, consortium U, Thomson W, Worthington J, Barton A. Study of the common genetic background for rheumatoid arthritis and systemic lupus erythematosus. Ann Rheum Dis. 2011;70:463–468. doi: 10.1136/ard.2010.137174. [DOI] [PMC free article] [PubMed] [Google Scholar] - 44.Qiu LJ, Ni J, Cen H, Wen PF, Zhang M, Liang Y, Pan HF, Mao C, Ye DQ. Relationship between the IL-4 gene promoter -590C/T (rs2243250) polymorphism and susceptibility to autoimmune diseases: a meta-analysis. J Eur Acad Dermatol Venereol. 2015;29:48–55. doi: 10.1111/jdv.12435. [DOI] [PubMed] [Google Scholar] - 45.Kumar A, Singh S, Goel F, Pandey RK, Singh L, Kumar A, Dobhal V. Molecular mechanisms and risk factors in rheumatoid arthritis: a comprehensive review. Inflammopharmacology. 2026;34:125–144. doi: 10.1007/s10787-025-02073-6. [DOI] [PubMed] [Google Scholar] - 46.Arnaud L, Chasset F, Martin T. Immunopathogenesis of systemic lupus erythematosus: an update. Autoimmun Rev. 2024;23:103648. doi: 10.1016/j.autrev.2024.103648. [DOI] [PubMed] [Google Scholar] - 47.Kumar M, Yip L, Wang F, Marty SE, Fathman CG. Autoimmune disease: genetic susceptibility, environmental triggers, and immune dysregulation. Where can we develop therapies? Front Immunol. 2025;16:1626082. doi: 10.3389/fimmu.2025.1626082. [DOI] [PMC free article] [PubMed] [Google Scholar] - 48.Amezcua-Guerra LM, Roldán-Ortega J, Mora-Ramírez M. Rhupus syndrome: current knowledge and future perspectives. Expert Rev Clin Immunol. 2025;21:1709–1724. doi: 10.1080/1744666X.2025.2605662. [DOI] [PubMed] [Google Scholar] - 49.Gray PE, David C. Inborn errors of immunity and autoimmune disease. J Allergy Clin Immunol Pract. 2023;11:1602–1622. doi: 10.1016/j.jaip.2023.04.018. [DOI] [PubMed] [Google Scholar] - 50.Sekula P, Del Greco MF, Pattaro C, Köttgen A. Mendelian randomization as an approach to assess causality using observational data. J Am Soc Nephrol. 2016;27:3253–3265. doi: 10.1681/ASN.2016010098. [DOI] [PMC free article] [PubMed] [Google Scholar] - 51.Liu M, Wang S, Liang Y, Fan Y, Wang W. Genetic polymorphisms in genes involved in the type I interferon system (STAT4 and IRF5): association with Asian SLE patients. Clin Rheumatol. 2024;43:2403–2416. doi: 10.1007/s10067-024-07046-8. [DOI] [PubMed] [Google Scholar] - 52.Barber MRW, Drenkard C, Falasinnu T, Hoi A, Mak A, Kow NY, Svenungsson E, Peterson J, Clarke AE, Ramsey-Goldman R. Global epidemiology of systemic lupus erythematosus. Nat Rev Rheumatol. 2021;17:515–532. doi: 10.1038/s41584-021-00668-1. [DOI] [PMC free article] [PubMed] [Google Scholar] - 53.Zhong H, Li XL, Li M, Hao LX, Chen RW, Xiang K, Qi XB, Ma RZ, Su B. Replicated associations of TNFAIP3, TNIP1 and ETS1 with systemic lupus erythematosus in a southwestern Chinese population. Arthritis Res Ther. 2011;13:R186. doi: 10.1186/ar3514. [DOI] [PMC free article] [PubMed] [Google Scholar]