Sage Journals: Discover world-class research

Abstract

Background:

Rheumatoid arthritis might increase the risk of lung diseases, such as interstitial lung disease and lung cancer. However, the causal relationships remained unclear. Herein, we utilized Mendelian randomization to study the causal relationships.

Material and methods:

Genetic Data from European ancestry and East Asian ancestry were used. The main Mendelian randomization approaches included inverse variance weighted, weighted median, and Mendelian randomization-Egger. We also conducted a series of analyses to validate the robustness of the Mendelian randomization estimates by integrating complementary Mendelian randomization models, Cochran’s Q test, Mendelian randomization-Egger intercept analysis, leave-one-out analysis, RadialMR, Steiger filtering test, multivariable Mendelian randomization and colocalization analysis.

Results:

In East Asian, inverse variance weighted indicated a significant association between rheumatoid arthritis and lung cancer after Bonferroni correction (odds ratio = 1.11, 95% confidence interval: 1.05–1.18); in both East Asian and European ancestries, rheumatoid arthritis was found to increase risk for interstitial lung disease (odds ratio_{East Asian} = 1.29, 95% confidence interval: 1.17–1.42, which remained significant after Bonferroni correction; odds ratio_European = 1.08, 95% confidence interval: 1.02–1.16). The observed significant associations between rheumatoid arthritis and interstitial lung disease remained in both ancestries after accounting for smoking behavior in multivariable Mendelian randomization analyses. The detected impact of rheumatoid arthritis on the risk of lung cancer in East Asian ancestry vanished after adjusted with smoking initiation.

Conclusion:

Using Mendelian randomization analysis, our results unveiled that rheumatoid arthritis augmented the risks of IL, independent of smoking initiation in both European and East Asian ancestry. These revelations might have implications for clinicians to be more vigilant and enhance routine interstitial lung disease screening in rheumatoid arthritis patients.

Keywords

rheumatoid arthritis lung diseases causality

Introduction

Rheumatoid arthritis (RA) is a chronic, systemic autoimmune inflammatory disorder.¹ It has a predilection for affecting women over the age of 40.² On a global scale, the prevalence of RA ranges from 0.25% to 1%.³ In recent years, due to the development of novel therapeutic approaches, significant progress has been made in the treatment and prognosis of RA. However, this has also shifted the focus toward the management of complications associated with the disease. Before effectively managing these complications, a comprehensive understanding of the complications that RA patients may develop is essential. Previous research has indicated that RA can have an impact on multiple systems, such as the cardiovascular system, diabetes, respiratory system, and nervous system, among others.⁴ Gaining a full understanding of the complications caused by RA can contribute to the formulation of more optimized treatment strategies for RA patients.

Both lung diseases and RA are prevalent conditions that can cause significant disability.^5,6 They share common pathogenic pathways, such as tobacco use and pulmonary inflammation. The influence of respiratory conditions, particularly chronic obstructive pulmonary disease (COPD), on RA has been observed. Conversely, RA may also be one of the risk factors for the development of COPD. A review has reported that RA increases the risk of COPD (odds ratio (OR) = 1.41, 95% confidence interval (95% CI): 1.13–1.76, I² = 97.8%; p = 0.003). On the other hand, it has also been reported that COPD increases the risk of RA (OR = 1.36, 95% CI: 1.05–1.76, I² = 55.0%; p = 0.022).⁷ However, the existence of a bidirectional association between RA and COPD remains unclear. Moreover, other lung diseases like asthma, interstitial lung disease (ILD), and lung cancer (LC) have also been suggested as factors that may influence RA.^8–11 Whether RA can lead to the development of LC and other lung diseases is still uncertain.

Mendelian randomization (MR) is a powerful analytical method for elucidating the causal relationship between an exposure and an outcome.¹² It uses genetic variants to infer exposure, which can effectively minimize confounding bias. In the past, MR has been widely applied in the discovery of disease biomarkers and complications. Indeed, with advances in MR techniques, there is growing research interest in investigating the underlying causal relationships between comorbid conditions. Thus, MR has become a key methodological tool that makes it feasible to investigate causal links between distinct diseases. For example, Ryszkiewicz et al. explored the causal effects of ADHD and autism on cardiovascular disease risk.¹³ Another study also reported the causal associations between type 2 diabetes and various cancers using MR techniques.¹⁴ It has successfully helped to identify the causal relationships between RA and various factors such as peripheral artery disease, osteoporosis, etc.^15,16 Nevertheless, to date, there has been no comprehensive MR analysis exploring the relationship between RA and lung diseases. Therefore, the aim of our study is to investigate the relationship between RA and common lung diseases, including asthma, COPD, ILD, and LC.

Methods

Study design

This study encompassed four outcomes, namely: (1) asthma; (2) COPD; (3) ILD; and (4) LC. Three MR methods were employed to examine the causal relationships between RA and these four outcomes in both Asian and European populations.¹⁷ The overall study framework is presented in Figure 1. For a convincing causal relationship to be established, it must adhere to the three fundamental MR assumptions: relevance, exchangeability, and exclusivity. Specifically, the instrumental variables (IVs) should exhibit a strong correlation with the exposure factor; the causal connection between the IVs and the outcomes should occur solely through the exposure factor; and the causal relationship between the exposure and the outcome should not be distorted by confounding factors. When all these assumptions are satisfied, MR can serve as a powerful tool for uncovering causal relationships. Our study was reported in accordance with the guidelines for MR reporting.¹⁸

Figure 1.

Study design of a Mendelian randomization study of rheumatoid arthritis on lung diseases.

GWAS data for the exposure

Regarding the East Asian population, the data on RA were obtained from the Biobank of Japan with the access code bbj-a-151. The data can be downloaded from the website of the Biobank of Japan as well as the IEU GWAS Catalog. In total, 212,453 participants were recruited, among whom 4199 were diagnosed with RA.

For the European population, the data on RA were sourced from the IEU GWAS Catalog with the access code ieu-a-832. Altogether, 58,284 participants were enrolled, and 14,361 of them were diagnosed with RA.

Selection of IVs

The selection of IVs constituted the most crucial step in the MR analysis. In this study, we strictly adhered to the standard MR protocol.

The IVs for RA were selected based on the following inclusion criteria: (i) p < 5 × 10⁻⁸ to identify relevant genetic instruments¹⁹; (ii) to preclude linkage disequilibrium, clustering was performed with r² < 0.001 and a distance of 10 MB was maintained between loci based on the European or Asian panel of 1000 Genome Project; (iii) the F statistics for each single nucleotide polymorphism (SNP) were calculated, and those with F < 10 were removed to mitigate weak instrument bias.²⁰ These criteria have been widely recognized and utilized in previous literature. Before formal MR analysis, harmonization and palindromic SNP handling were conducted. In detail, the exposure SNPs were extracted from the outcome datasets and then harmonization was carried out to align the effect alleles and exclude incompatible SNPs or palindromic SNPs with intermediate minor allele frequencies (MAF) (MAF > 0.42).

GWAS data for the outcomes

In this study, four lung diseases were analyzed, namely asthma, COPD, ILD, and LC. For the Asian population, the data for these four lung diseases were obtained from ebi-a-GCST90018575, ebi-a-GCST90018587, ebi-a-GCST90018643, and ebi-a-GCST90018655, respectively. The disease endpoints were defined based on the International Classification of Diseases-10 (ICD-10) classifications (https://icd.who.int/browse10/2019/en).

For the European population, the data for these four lung diseases were derived from ebi-a-GCST006862, ukb-b-20464, ebi-a-GCST90018863, and ebi-a-GCST004748, respectively. Among them, diagnosis of asthma was based on standardized questionnaires or physicians’ evaluation; COPD was defined by doctors’ diagnosis; ILD was defined based on ICD-10; and LC was diagnosed histologically based on doctors. In addition, considering the availability of subtypes for LC phenotype in European ancestry, we additionally evaluated the impact of RA on small cell LC (SCLC; “finn-b-C3_SCLC_EXALLC”) and non-SCLC (NSCLC; “finn-b-C3_LUNG_NONSMALL”).

These samples provided the fundamental data support for subsequent investigations into the relationships between RA and various outcome measures. A sufficient sample size contributed to enhancing the reliability and representativeness of the research results. Notably, there was no sample overlap between the exposure and outcome datasets from European ancestry. However, the exposure and outcome data for Asian ancestry were primarily derived from the Japanese biobank, which might induce false-positive estimates. Therefore, we performed maximum likelihood estimation to verify the observed significant associations from East Asian ancestry.

MR analysis and sensitivity analysis

To obtain reliable estimates, we utilized three MR methods, including inverse variance weighted (IVW)²¹ R-Egger, and weighted median (WM).²²

Both IVW and WM were regarded as the primary analysis methods. The MR-Egger method was used as a supplementary approach to test whether pleiotropic outliers would bias the direction of the IVW and WM estimates. The IVW method is the most basic, assuming that all IVs are valid. It weights each genetic association by the inverse of its variance, yielding more reliable estimates when most IVs are valid. In this work, random-effect IVW was applied when heterogeneity was detected. The WM method can produce consistent estimates even when up to 50% of the IVs are invalid, by using a WM of the individual-SNP estimates. The MR-Egger method addresses the potential pleiotropy issue. It can account for directional pleiotropy by estimating and adjusting for the intercept in a regression model, permitting valid causal inference in the presence of some invalid IVs. In this work, causal estimation was considered statistically significant with p < 2.08 × 10⁻³ (Bonferroni-corrected p = 0.05/two ancestries/four outcomes/three major estimators). Statistical power was calculated for the estimates derived from the IVW method using an online tool (https://shiny.cnsgenomics.com/mRnd/).

In addition, sensitivity analyses were also conducted, including additional complementary MR approaches (MR-PRESSO, MR-RAPS, and Contamination Mixture) aside from the above three conventional MR models, Cochran’s Q test, MR intercept test, leave-one-out analysis, and Steiger filtering test. These methods primarily focused on heterogeneity and pleiotropy. Cochran’s Q test was used to detect heterogeneity among the effect estimates from different IVs. High heterogeneity could imply that the causal relationships estimated by different genetic variants were inconsistent, which might be due to unaccounted confounding or pleiotropic effects. The MR intercept test, often implemented in the MR-Egger regression, was crucial for identifying pleiotropy. If the intercept was significantly different from zero, it suggested the presence of directional pleiotropy, meaning that the IVs were associated with the outcome through other pathways besides the exposure of interest, potentially biasing the causal estimates. Leave-one-out analysis was a sensitivity analysis in MR. By successively removing each genetic variant (IV) and re-calculating the causal estimate, it helped to determine if any single variant was driving the overall result. If the estimate changed substantially when a particular variant was removed, that variant might be an outlier or have a strong influence on the outcome. We also evaluated the direction of the MR analysis using the Steiger direction test. Specifically, we conducted Steiger filtering for the detected causalities from RA to lung diseases.

If heterogeneity existed, RadialMR would be carried out to detect outliers.²³ It was designed to estimate causal effects using genetic variants. It focused on the radial structure of data in the context of multiple IVs. By leveraging the relationships between genetic instruments, exposure, and outcome, RadialMR aimed to provide more accurate and robust causal inferences. It could handle complex data structures and was useful in various fields like epidemiology and genetics. This method helped to better understand the causal links between traits and diseases, offering valuable insights for research and potential interventions. A new estimate after removing outliers would be regarded as the final estimate.

Excluding the potential bias of smoking behavior

Considering lung diseases are largely associated with tobacco intake, smoking phenotype should be considered when investigating the causal associations between RA and lung diseases. To this end, this work conducted multivariable MR (MVMR) analysis of RA with lung diseases adjusted with smoking initiation (“ieu-b-4877” for European ancestry and “bbj-a-78” for East Asian ancestry).

In addition, we manually research whether the RA exposure SNPs were associated with several GWAS traits that might influence the true associations between RA and lung diseases, including smoking behavior, forced vital capacity (FVC), and forced expiratory volume in 1 s (FEV1), mainly from Pan-UKB datasets.

Bayesian colocalization analysis

Bayesian colocalization analysis tests the posterior probability for five competing hypotheses (H0–H4) within a defined genomic region to determine if the exposure and outcome traits share a single common causal variant. The hypotheses are: H0: no association with either trait; H1: association with the exposure only; H2: association with the outcome only; H3: association with both traits, driven by two distinct causal variants; H4: association with both traits, driven by a single shared causal variant. We extracted effect size estimates and their standard errors for the exposure and outcome-associated SNPs from the respective GWAS summary statistics for the genomic regions of interest. A posterior probability for H4 (PP.H4) ⩾ 80% was considered strong evidence for colocalization, indicating that the genetic association signals for both traits are likely attributable to the same underlying causal genetic variant. In this work, we conducted colocalization using a window of 100 kb surrounding the top-SNP of RA datasets from East Asian and European ancestry.

Reverse MR analysis

Except for investing the causalities from RA to common lung diseases, this work also conducted a reverse MR analysis to evaluate the causal impact of common lung diseases on RA. For East Asian ancestry, we evaluated the effects of asthma, COPD, LC on RA using a p threshold of 5 × 10⁻⁸ exposure-SNP, except for ILD using 5 × 10⁻⁶ due to limited SNPs available at a threshold of genome-wide significance. For European ancestry, we evaluated the effects of asthma and LC on RA using a p threshold of 5 × 10⁻⁸ exposure-SNP, except of ILD (5 × 10⁻⁷) and COPD (5 × 10⁻⁶) due to limited SNPs available at the threshold of genome-wide significance.

Statistical analysis

The analysis in this study relied on the R platform (version 4.3.2, 64-bit versions of Windows 7 and later on x86_64 chips, developed at Bell Laboratories: https://www.r-project.org/). Packages “TwoSampleMR,” “MendelianRandomization,” “RadialMR,” “forestplot,” “coloc,” and “data.table” were used in our study to conduct MR analysis and visualization.

Results

Instruments overview

The included exposure-instruments with genetic information, including sample sizes, R² (variance of explanation per single SNP), and F-statistics were presented in Tables S1 and S2. There were no weak instruments included for MR analysis as F-statistics for each of the SNPs were all over 10. The harmonized data for MVMR analysis was presented in Tables S3 to S5.

Causal relationship between RA and lung diseases in East Asian ancestry

In the context of East Asian ancestry, evidence derived from the IVW method indicated that RA augmented the risk of ILD and LC (Table 1). The IVW approach suggested that the OR for ILD in RA patients relative to non-RA patients was 1.29 (95% CI: 1.17–1.42; p = 1.72 × 10⁻⁷; Figure 2). The detected association was statistically significant with even after Bonferroni correction (p < 2.08 × 10⁻³). The WM and MR-Egger methods yielded comparable estimates and remained significant, with values of 1.27 (95% CI: 1.14–1.42) and 1.22 (95% CI: 1.04–1.42), respectively. Neither heterogeneity nor pleiotropy was observed in the MR analysis between RA and ILD (p = 0.36 and p = 0.39, respectively). In addition, maximum likelihood analysis showed the observed association remained significant (OR = 1.27, 95% CI: 1.15–1.41; p = 4.3 × 10⁻⁶). The WM estimates suggested that RA increased the risk for LC (OR = 1.09, 95% CI: 1.03–1.51; p = 3.51 × 10⁻³) and reduced the risk for asthma (OR = 0.93, 95% CI: 0.88–0.97; p = 0.03). However, the maximum likelihood analysis showed the association between RA and LC vanished (OR = 1.09, 95% CI: 0.99–1.09; p = 0.09), indicating the observed association in prior analyses might be biased by sample overlapping. There was no established causal relationship between RA and either asthma or COPD. In the MR analysis between RA and LC, no pleiotropy was detected (p = 0.86), yet heterogeneity was present (p = 2.37 × 10⁻⁵). Marked heterogeneity was evident in the MR analysis between RA and asthma (p = 1 × 10⁻¹³). Steiger filter test showed that all the SNPs explained more variance in the exposure (RA) than in the outcomes (lung diseases; Table S1).

Table 1.

MR estimates the association between rheumatoid arthritis and lung diseases.

Exposure	Outcomes	Method^a	nSNP	OR	p	Power	Heterogeneity			Pleiotropy
Exposure	Outcomes	Method^a	nSNP	OR	p	Power	Q	I² (%)	Q–p	Intercept	p
Rheumatoid arthritis (East Asia)	Asthma^b	IVW	10	0.98 (0.90–1.07)	0.69	5%	78.45	85.98	1.00E-13	0.02	0.56
		Weighted median	10	0.93 (0.88–0.97)	0.003
		MR-Egger	10	0.95 (0.83–1.09)	0.50
	Asthma^c	IVW	6	1.02 (0.90–1.15)	0.80	–	13.28	52.71	0.02
		Weighted median	6	1.07 (0.96–1.19)	0.24
		MR-Egger	6	1.33 (0.77–2.28)	0.36
	COPD	IVW	10	1.01 (0.95–1.08)	0.73	5%	16.22	32.18	0.06	0.04
		Weighted median	10	1.00 (0.94–1.07)	0.90
		MR-Egger	10	0.94 (0.86–1.02)	0.17
	ILD	IVW^d	10	1.29 (1.17–1.42)	1.72E-07	8%	9.86	0	0.36	0.03	0.39
		Weighted median^d	10	1.27 (1.14–1.42)	1.35E-05
		MR-Egger	10	1.22 (1.04–1.42)	0.04
	LC^b	IVW	10	1.10 (1.00–1.20)	0.046	7%	37.25	70.47	2.37E-05	0.01	0.86
		Weighted median	10	1.09 (1.03–1.15)	0.002
		MR-Egger	10	1.08 (0.93–1.27)	0.34
	LC^c	IVW^d	6	1..11 (1.05–1.18)	4.18E-04	–	6.12	16.34	0.29
		Weighted median	6	1.09 (1.03–1.51)	3.51E-03
		MR-Egger	6	1..06 (0.98–1.15)	0.22
Rheumatoid arthritis (Europe)	Asthma	IVW	16	0.96 (0.82–1.12)	0.59	12%	180.25	90.57	2.11E-30	−0.01	0.6
		Weighted median	16	0.97 (0.92–1.04)	0.41
		MR-Egger	16	1.03 (0.76–1.38)	0.86
	COPD	IVW	26	1.00 (0.998–1.002)	0.94	5%	23.34	0	0.56	0.0005	0.21
		Weighted median	26	1.00 (0.997–1.002)	0.60
		MR-Egger	26	0.995 (0.987–1.003)	0.23
	ILD	IVW	39	1.08 (1.02–1.16)	0.01	23%	68.36	41.49	1.81E-03	0.01	0.51
		Weighted median	39	1.04 (0.97–1.12)	0.24
		MR-Egger	39	1.06 (0.96–1.17)	0.27
	LC	IVW	36	0.98 (0.95–1.02)	0.37	12%	78.08	52.61	3.97E-05	−0.01	0.07
		Weighted median	36	1.01 (0.97–1.04)	0.70
		MR-Egger	36	1.02 (0.97–1.08)	0.41

COPD: chronic obstructive pulmonary disease; ILD: interstitial lung disease; IVW: inverse variance weighted; LC: lung cancer; MR: Mendelian randomization.

The IVW method was based on the random-effect model to balance potential heterogeneity.

Pre-RadialMR exploratory analysis showing significant heterogeneity with the weighted median method showing significant estimates.

MR estimates after outliers removed by RadialMR analysis.

Statistically significant estimate after Bonferroni p correction.

Figure 2.

Forest plot showing Mendelian randomization estimates of rheumatoid arthritis on lung diseases.

To mitigate the influence exerted by heterogeneity, we employed the RadialMR method. In the MR analysis between RA and LC, outliers were identified at rs12612769, rs1557549, rs1634734, and rs9275610. After excluding these outliers, a causal relationship between RA and LC was established. After Bonferroni correction, the IVW method indicated a significant association between RA and the risk of LC (OR = 1.11, 95% CI: 1.05–1.18; p = 4.18 × 10⁻⁴). The WM approach furnished a similar estimate and retained significance (OR = 1.09, 95% CI: 1.03–1.51), whereas the MR-Egger estimate exhibited a consistent direction but was non-significant (OR = 1.06, 95% CI: 0.98–1.15). Heterogeneity was largely mitigated (Q = 6.12, Q–p = 0.29, I² = 16.34). For the MR analysis between RA and asthma, outliers were detected at rs117530403, rs1557549, rs1634734, and rs9275610. Upon removing these outliers, the IVW estimates remained non-significant (OR = 1.02, 95% CI: 0.90–1.15; p = 0.80). Similarly, we re-conducted Cochran’s Q test and the results showed heterogeneity was largely mitigated (Q = 13.28, Q–p = 0.02, I² = 52.71%). In addition, complementary MR estimations, including MR-PRESSO, RAPS, and Contamination Mixture, presented similar results (Table S6). With a relatively modest sample size in East Asian, the statistical power was limited for asthma (5%), COPD (5%), ILD (8%), and LC (7%; Table 1).

Causal relationship between RA and lung diseases in European ancestry

In the European ancestry cohort, the IVW estimate provided support for the notion that RA increased the risk of ILD with nominal significance (OR = 1.08, 95% CI: 1.02–1.16; p = 0.01). No pleiotropy was detected (p = 0.51), although heterogeneity was present. No causal relationships were discerned between RA and asthma, COPD, or LC (including NSCLC and SCLC) in the European ancestry population (Table S7). The scatter plots and leave-one-out plots of MR analysis between RA and ILD in East Asian and European ancestry were presented in Figures 3 and 4. The Steiger filter test showed that all the SNPs explained more variance in the exposure (RA) than in the outcomes (lung diseases), except for the SNP rs12232497 from the nonsignificant association between RA and asthma (Table S2). However, leave-one-out analysis showed no evidence of this variant’s influence. Similarly, modest statistical power was observed in the associations derived from European ancestry (12% for asthma, 5% for COPD, 23% for ILD, and 12% for LC).

Figure 3.

Scatter plots showing Mendelian randomization estimates of rheumatoid arthritis on interstitial lung disease in East Asian ancestry (a) and European ancestry (b).

Figure 4.

Leave-one-out plots showing Mendelian randomization estimates of rheumatoid arthritis on interstitial lung disease in East Asian ancestry (a) and European ancestry (b).

Excluding the influence of smoking

For East Asian ancestry, MVMR analysis showed that RA was significantly associated with ILD when adjusting for smoking initiation (OR = 1.29, 95% CI: 1.17–1.43; p = 1.04 × 10⁻⁶). However, the detected association between RA and LC vanished when taking smoking into account (OR = 1.07, 95% CI: 0.99–1.16; p = 0.61). In European ancestry, MVMR analysis indicated that RA had a detrimental effect on ILD independent of smoking initiation (OR = 1.11, 95% CI: 1.04–1.19; p = 0.002).

In addition, SNP pleiotropy research, mainly based on Pan-UKB datasets, showed that the RA-exposure SNPs were not associated with smoking, FEV1, or FVC (Table S8).

Colocalization analysis

Based on the two-sample MR analyses, sensitivity analyses and MVMR analyses, we identified that RA was stably associated with ILD in both ancestries. Further colocalization analysis confirmed that these causalities were not driven by common SNPs (PP.H4 = 0.43 for the East Asian ancestry; PP.H4 = 0.01 for the European ancestry; Table S9).

Reverse MR analyses

Harmonized data for reverse MR analysis was presented in Table S10 and S11. For the East Asian analysis, we did not find evidence of causal relationships from four lung diseases to RA (Table S12). While for the European ancestry, reverse MR analysis showed that ILD was associated with an increased risk of RA (OR = 1.36, 95% CI: 1.17–1.58; p = 5.98 × 10⁻⁵). No evidence was found for the causal effect of asthma, LC, or COPD on RA in the European ancestry (Table S13).

Discussion

To the best of our knowledge, this study represents the maiden MR analysis exploring the causal associations between RA and four pulmonary diseases. Our results unveiled that RA augmented the risks of IL independent of smoking initiation in both European and East Asian ancestry. These revelations might have implications for clinicians to be more vigilant and enhance routine ILD screening in RA patients. In addition, the reverse MR analysis also indicated a causal impact of ILD on the risk of RA in European ancestry.

In recent years, the management paradigm for RA has experienced a paradigm shift. It has transitioned from a straightforward translation of clinical trial efficacy to patient care, now encompassing not only the management of RA but also a holistic consideration of other health aspects. Notably, with the successful application of disease-modifying antirheumatic drugs in RA treatment, the focus has broadened to overall patient well-being.²⁴ Consequently, our study, which aimed to decipher the causal links between RA and lung diseases, holds profound clinical significance.

ILD stood as a prevalent yet often underappreciated complication of RA, bearing significant morbidity and mortality implications.^25–30 ILD can manifest at any stage during the course of RA and was often diagnosed with high-resolution computed tomography (HRCT) or histological examination.³¹ The clinical trajectory of RA-ILD is highly variable.^32,33 Close surveillance of ILD in RA patients was vital to ensure timely intervention with appropriate treatment strategies, thereby improving outcomes. Clinical trials have demonstrated that antifibrotic therapies can slow the decline in lung function in patients with progressive fibrosing ILDs. The management of RA-ILD patients necessitated a multidisciplinary approach, evaluating both the severity and progression of ILD and the activity of articular disease. Close cooperation between rheumatologists and pulmonologists was indispensable for optimizing patient care.

In summary, RA and ILD, LC shared the following pathogenesis- inflammatory response, immune dysregulation, and shared genetic susceptibility.

A primary mechanism potentially linking RA to ILD and LC is the chronic inflammatory state intrinsic to RA. In RA, the immune system is in a state of persistent activation, leading to the release of a profusion of pro-inflammatory cytokines such as tumor necrosis factor-alpha (TNF-α), interleukin-1 (IL-1), and interleukin-6 (IL-6). These cytokines can exert a profound influence on the lung microenvironment.

In the context of ILD, the continuous exposure of lung tissue to these pro-inflammatory cytokines can inflict damage on the alveolar epithelium and the pulmonary interstitium. For instance, TNF-α can induce apoptosis of alveolar epithelial cells, disrupt the normal architecture of the alveolar-capillary barrier, and promote fibroblast activation and proliferation. This fibroblast activation culminates in the excessive production of extracellular matrix components, ultimately giving rise to pulmonary fibrosis, a hallmark of interstitial pneumonia.^34–37

Regarding LC, chronic inflammation can foster a tumor-promoting microenvironment.^38,39 Inflammatory cells infiltrating the lung tissue release reactive oxygen species and reactive nitrogen species as part of the immune response. These highly reactive molecules can inflict DNA damage on normal lung cells, triggering mutations that may initiate the carcinogenic process. Moreover, pro-inflammatory cytokines can promote cell proliferation, angiogenesis, and inhibit apoptosis in transformed cells, facilitating the growth and metastasis of LC cells.

RA is accompanied by immune dysregulation, which may also contribute to the elevated risks of interstitial pneumonia and LC. In RA patients, there is an abnormal activation of T-lymphocytes and B-lymphocytes. The dysregulated T-cell subsets can secrete cytokines that disrupt the normal immune surveillance in the lungs.

For interstitial pneumonia, autoreactive T-cells may erroneously target the lung tissue, perceiving it as foreign. This autoimmune assault can lead to the destruction of lung parenchyma and the development of fibrosis. Additionally, the imbalance in B-cell function can result in the production of autoantibodies that deposit in the lung tissue, triggering an inflammatory response and further tissue damage.

In the context of LC, the impaired immune surveillance due to immune dysregulation in RA may permit cancer cells to elude the immune system.^40,41 The normal immune response is designed to recognize and eliminate transformed cells. However, in RA patients, the abnormal immune cell activation and cytokine production can interfere with this process. For example, regulatory T-cells, which are over-activated in RA, can suppress the anti-tumor immune response, allowing cancer cells to proliferate and metastasize without effective control.

There may exist a shared genetic susceptibility among RA, interstitial pneumonia, and LC. Some genetic variants associated with RA may also modulate the susceptibility to lung diseases. For example, certain major histocompatibility complex alleles linked to RA may also participate in the immune response to lung-specific antigens. These genetic factors can impact the function of immune cells, cytokine production, and the regulation of inflammation in the lungs.

Genetic variants that heighten the risk of RA may also disrupt the normal repair mechanisms in the lungs. In the case of interstitial pneumonia, impaired repair of damaged lung tissue due to genetic factors can fuel the progression of fibrosis. For LC, genetic mutations associated with both RA and an increased risk of carcinogenesis may be implicated in the initiation and development of lung tumors.

Future research is warranted to comprehensively elucidate these mechanisms, which will lay the foundation for the development of more efficacious preventive and treatment strategies for these comorbid conditions.

Previously, several studies posited that RA might elevate the risk of COPD. However, our MR analysis yielded the finding that RA did not influence the occurrence of COPD in either Asian or European ancestry. Hence, there is no imperative for special preventative measures against COPD in RA patients.

Notably, though we conducted a series of analyses (including complementary MR models, sensitivity analyses, MVMR analyses, manually researching pleiotropic effects of SNPs, and Steiger direction test) to confirm the stable associations between RA and ILD, colocalization analysis failed to detect common variants driving the causalities. There might be multiple independent causal signals (allelic heterogeneity) for the exposure and the outcome. Colocalization methods typically model a single causal variant per region. When this assumption is violated, the power of colocalization analysis decreases, making it less likely to support a shared causal variant (H4), even if a causal relationship exists at the phenotypic level captured by MR. Crucially, the validity of our MR finding does not strictly depend on a positive colocalization result. Our MR analysis was based on independent, genome-wide significant instruments and was robust across multiple sensitivity methods that are resistant to certain forms of pleiotropy. This provides a consistent and robust evidence for a causal effect, albeit one with a more complex underlying genetic basis than a single shared variant.

There were several limitations of this study. Though MR provided insight into causal relationships, this study did not integrate clinical trials. Though this study revealed causal relationships in two ancestries, it did not subgroup patients into different sexes, different disease stages, etc. Future work is warranted to investigate the sex or age stratified MR estimates when corresponding genome-wide association studies (GWAS) data was available. In addition, power calculation was modest for most MR estimates in this work, which might be due to limited variance explained by SNPs or modest sample sizes in this work. Future work is warranted to validate our results using larger GWAS cohorts. Besides, we conducted MVMR analysis to correct smoking behavior by including only the smoking initiation phenotype, which was obtained from a large-scale GWAS cohort. Other smoking-related phenotypes, like cigarettes per day or lifetime smoking, were not included due to modest sample size or data use restriction. Finally, while MR benefits from inherent methodological strengths that substantially reduce confounding and provide robust evidence for causal inference between phenotypes, its findings still warrant cautious interpretation in clinical contexts, which has also been highlighted in Cai et al.’s study.⁴² Indeed, as a robust tool for causal inference, MR is widely used to preliminarily explore the causal relationships between different phenotypes, and its preliminary conclusions might guide further real-world research. While our multi-method MR analyses support a causal relationship between RA and lung diseases and may offer clinically relevant insights, further validation through prospective cohort studies is needed to confirm these associations.

Conclusion

Using MR analysis, RA would increase the risk for ILD in both East Asian ancestry and European ancestry even after accounting for smoking behavior. Though the primary two-sample MR analysis detected a detrimental impact of RA on risk of LC, this effect vanished after adjusted with smoking behavior, indicating the observed unfavorable impact of RA on LC was facilitated by smoking behavior. These revelations might have implications for clinicians to be more vigilant and enhance routine ILD screening in RA patients.

Supplemental Material

sj-xlsx-1-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-1-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-10-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-10-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-11-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-11-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-12-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-12-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-2-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-2-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-3-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-3-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-4-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-4-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-5-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-5-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-6-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-6-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-7-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-7-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-8-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-8-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Supplemental Material

sj-xlsx-9-smo-10.1177_20503121261420918 – Supplemental material for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study

Supplemental material, sj-xlsx-9-smo-10.1177_20503121261420918 for Rheumatoid arthritis and risk of lung diseases in East Asians and Europeans: Evidence from a Mendelian randomization study by Shengkai Zhang, Xiang Huang, Meiying Huang, Xinhan Peng, Qiuye Cao and Yi Lan in SAGE Open Medicine

Footnotes

Acknowledgements

We would like to thank the participants join the GWAS analysis utilized in our study.

ORCID iD

Shengkai Zhang

Author contributions

S.Z. and Y.L. designed the study; S.Z. and X.H. conducted major statistical analyses; M.H. and X.P. conducted visualization; S.Z. wrote the first version of the article. All the authors reviewed and approved the final version of the article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The datasets for this study were publicly available and could be found from the IEU consortium.

Supplemental material

Supplemental material for this article is available online.

References

Di Matteo

Bathon

Emery

. Rheumatoid arthritis. Lancet 2023; 402(10416): 2019–2033.

Scott

Whittle

Bailey

, et al. Rheumatoid arthritis, psoriatic arthritis, and axial spondyloarthritis epidemiology in England from 2004 to 2020: an observational study using primary care electronic health record data. Lancet Reg Health Eur 2022; 23: 100519.

Finckh

Gilbert

Hodkinson

, et al. Global epidemiology of rheumatoid arthritis. Nat Rev Rheumatol 2022; 18: 591–602.

Kattamuri

Duggal

Aparece

, et al. Cardiovascular risk factor and atherosclerosis in rheumatoid arthritis (RA). Curr Cardiol Rep 2025; 27(1): 31.

Ruiz-Esquide

Gómara

Peinado

, et al. Anti-citrullinated peptide antibodies in the serum of heavy smokers without rheumatoid arthritis. A differential effect of chronic obstructive pulmonary disease? Clin Rheumatol 2012; 31: 1047–1050.

Scherer

Häupl

Burmester

GR.

The etiology of rheumatoid arthritis. J Autoimmun 2020; 110: 102400.

Wang

Pan

Zhai

, et al. Bidirectional association between rheumatoid arthritis and chronic obstructive pulmonary disease: a systematic review and meta-analysis. Front. Immunol 2024; 15: 1494003.

Charoenngam

Ponvilawan

Rittiphairoj

, et al. Patients with asthma have a higher risk of rheumatoid arthritis: a systematic review and meta-analysis. Semin Arthritis Rheum 2020; 50: 968–976.

Ford

Liu

Chu

, et al. Asthma, chronic obstructive pulmonary disease, and subsequent risk for incident rheumatoid arthritis among women: a prospective cohort study. Arthritis Rheumatol 2020; 72: 704–713.

10.

Akiyama

Kaneko

Pathogenesis, clinical features, and treatment strategy for rheumatoid arthritis-associated interstitial lung disease. Autoimmun Rev 2022; 21: 103056.

11.

Kronzer

Hayashi

Crowson

, et al. Gene-respiratory disease interactions for rheumatoid arthritis risk. Semin Arthritis Rheum 2023; 63: 152254.

12.

Davies

Holmes

Davey Smith

Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ 2018; 362: k601.

13.

Ryszkiewicz

Malinowska

Jasińska-Stroschein

Evaluating the causal effects of ADHD and autism on cardiovascular diseases and vice versa: a systematic review and meta-analysis of Mendelian randomization studies. Cells 2025; 14(15): 1180.

14.

Pearson-Stuttard

Papadimitriou

Markozannes

, et al. Type 2 diabetes and cancer: an umbrella review of observational and Mendelian randomization studies. Cancer Epidemiol Biomarkers Prev 2021; 30(6): 1218–1228.

15.

Zheng

Zhou

, et al. Genetic association between immune-mediated inflammatory diseases and peripheral artery disease: a Mendelian randomization study. Sci Rep 2025; 15(1): 3891.

16.

Pan

Chen

, et al. Mendelian randomization studies of risk and protective factors for osteoporosis: a systematic review and meta-analysis. Front Endocrinol 2025; 15: 1486188.

17.

Chen

Kong

Pan

, et al. Kidney damage causally affects the brain cortical structure: a Mendelian randomization study. EBioMedicine 2021; 72: 103592.

18.

Skrivankova

Richmond

Woolf

BAR

, et al. Strengthening the Reporting of Observational Studies in Epidemiology using Mendelian randomisation (STROBE-MR): explanation and elaboration. BMJ 2021; 375: n2233.

19.

Fadista

Manning

Florez

, et al. The (in)famous GWAS P-value threshold revisited and updated for low-frequency variants. Eur J Hum Genet 2016; 24(8): 1202–1205.

20.

Burgess

Thompson

SG.

Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol 2011; 40(3): 755–764.

21.

Burgess

Small

Thompson

SG.

A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res 2017; 26(5): 2333–2355.

22.

Bowden

Davey Smith

Haycock

, et al. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol 2016; 40(4): 304–314.

23.

Bowden

Spiller

Del Greco

, et al. Improving the visualization, interpretation and analysis of two-sample summary data Mendelian randomization via the radial plot and radial regression. Int J Epidemiol 2018; 47: 1264–1278.

24.

Konzett

Aletaha

Management strategies in rheumatoid arthritis. Nat Rev Rheumatol 2024; 20(12): 760–769.

25.

Bongartz

Nannini

Medina-Velasquez

, et al. Incidence and mortality of interstitial lung disease in rheumatoid arthritis: a population-based study. Arthritis Rheum 2010; 62: 1583–1591.

26.

Olson

Swigris

Sprunger

, et al. Rheumatoid arthritis interstitial lung disease-associated mortality. Am J Respir Crit Care Med 2011; 183: 372–378.

27.

Hyldgaard

Hilberg

Pedersen

, et al. A population-based cohort study of rheumatoid arthritis-associated interstitial lung disease: comorbidity and mortality. Ann Rheum Dis 2017; 76: 1700–1706.

28.

Zamora-Legoff

Krause

Crowson

, et al. Patterns of interstitial lung disease and mortality in rheumatoid arthritis. Rheumatology 2017; 56: 344–350.

29.

Raimundo

Solomon

Olson

, et al. Rheumatoid arthritis interstitial lung disease in the United States: prevalence, incidence, and healthcare costs and mortality. J Rheumatol 2019; 46: 360–369.

30.

Samhouri

Vassallo

Achenbach

, et al. The incidence, risk factors, and mortality of clinical and subclinical rheumatoid arthritis associated interstitial lung disease: a population-based cohort. Arthritis Care Res 2022; 74: 2042–2049.

31.

Lee

Kim

Yoo

, et al. Histopathologic pattern and clinical features of rheumatoid arthritis-associated interstitial lung disease. Chest 2005; 127: 2019–2027.

32.

Song

Lee

, et al. Clinical course and outcome of rheumatoid arthritis-related usual interstitial pneumonia. Sarcoidosis Vasc Diffuse Lung Dis 2013; 30: 103–112.

33.

Hyldgaard

Ellingsen

Hilberg

, et al. Rheumatoid arthritis associated interstitial lung disease: clinical characteristics and predictors of mortality. Respiration 2019; 98: 455–460.

34.

Zhang

Wang

, et al. Profibrotic effect of IL-17A and elevated IL-17RA in idiopathic pulmonary fibrosis and rheumatoid arthritis-associated lung disease support a direct role for IL-17A/IL-17RA in human fibrotic interstitial lung disease. Am J Physiol Lung Cell Mol Physiol 2019; 316: L487–L497.

35.

Paulin

Doyle

Fletcher

, et al. Rheumatoid arthritis-associated interstitial lung disease and idiopathic pulmonary fibrosis: shared mechanistic and phenotypic traits suggest overlapping disease mechanisms. Rev Invest Clin 2015; 67: 280–286.

36.

McInnes

Schett

The pathogenesis of rheumatoid arthritis. N Engl J Med 2011; 365: 2205–2219.

37.

Derksen

VFAM

Huizinga

TWJ

van der Woude

. The role of autoantibodies in the pathophysiology of rheumatoid arthritis. Semin Immunopathol 2017; 39: 437–446.

38.

England

Campany

Sayles

, et al. Associations of serum cytokines and chemokines with the risk of incident cancer in a prospective rheumatoid arthritis cohort. Int Immunopharmacol 2021; 97: 107719.

39.

England

Sokolove

Robinson

, et al. Associations of circulating cytokines and chemokines with cancer mortality in men with rheumatoid arthritis. Arthritis Rheumatol 2016; 68(10): 2394–2402.

40.

Turesson

Matteson

EL.

Malignancy as a comorbidity in rheumatic diseases. Rheumatology 2013; 52: 5–14.

41.

Szekanecz

Bakó

, et al. Malignancies in autoimmune rheumatic diseases – a mini-review. Gerontologist 2011; 57: 3–10.

42.

Cai

Wei

Chen

, et al. Socioeconomic status, individual behaviors and risk for mental disorders: a Mendelian randomization study. Eur Psychiatry 2022; 65(1): e28.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB

0.02 MB

0.04 MB

0.01 MB

0.04 MB

0.02 MB

0.01 MB