Rheumatoid arthritis (RA) patients face a two-fold increased risk of coronary artery disease compared to the general population, yet the underlying mechanisms remain poorly understood. We hypothesized that circulating macrophages serve as critical cellular mediators that infiltrate coronary arteries and drive atherosclerotic progression in RA patients. This study aims to elucidate the causal relationship between RA and coronary atherosclerosis and to identify key macrophage subpopulations and molecular pathways involved.

We employed a comprehensive integrative multi-omics approach combining multiple analytical strategies. Mendelian randomization (MR) analysis was performed using large-scale genome-wide association study (GWAS) data to establish causal inference. Single-cell RNA sequencing (scRNA-seq) of coronary artery specimens was conducted to characterize cellular heterogeneity at high resolution. Pseudotime trajectory reconstruction was applied to delineate macrophage differentiation dynamics. Cell-cell communication networks were inferred via CellChat to map intercellular signaling interactions. Hierarchical weighted gene co-expression network analysis (hdWGCNA) was utilized to identify co-expressed gene modules. Additionally, eight distinct machine learning algorithms were implemented for predictive modeling, with SHAP analysis providing model interpretability.

 MR analysis demonstrated causal relationships between 45 RA-associated genetic variants and coronary atherosclerosis risk, with key mediators including LPAR1 (OR=1.026, P=0.004) and NAAA (OR=1.025, P=0.009). Single-cell analysis identified 11 distinct macrophage subpopulations, among which Macrophages_NEW5 was specifically enriched in ribosomal protein genes (RPL and RPS families). hdWGCNA revealed co-expression modules dominated by translational machinery components, while pseudotime analysis demonstrated macrophage differentiation trajectories correlating with atherosclerotic progression. Cell-cell communication analysis unveiled extensive signaling networks involving MIF, COLLAGEN, and CD99 pathways between macrophages and vascular cells. Among eight machine learning models, logistic regression achieved the highest performance (AUC=0.941), and SHAP analysis identified EEF1A1, RPL18, and HNRNPA1 as the most predictive features. Non-negative matrix factorization further revealed distinct immune infiltration patterns associated with the ribosomal protein-enriched macrophage subset.

Our findings provide comprehensive evidence for macrophage-mediated mechanisms linking RA to coronary atherosclerosis, highlighting ribosomal protein-enriched macrophage subpopulations as potential therapeutic targets for cardiovascular risk reduction in RA patients. These results offer new insights into the pathogenesis of RA-associated cardiovascular comorbidity and may inform future biomarker development and targeted therapeutic strategies.