Exploratory Statistical Analysis of Rheumatoid Factor–Based Subgroups in Sjögren’s Syndrome

Rheumatoid factor (RF) is a core diagnostic marker for rheumatoid arthritis (RA) but is also common in Sjögren’s syndrome (SS). RF-positive SS patients often exhibit RA-like features, suggesting a potential overlapping subtype. This study investigated whether RF status in SS is associated with distinct autoantibody profiles and greater immunological similarity to RA. We analyzed a synthetic dataset (n=176), including 91 RA patients and 85 SS patients (38 RF-positive, 45 RF-negative), across four autoantibodies: anti-citrullinated protein antibodies (ACPA), antinuclear antibodies (ANA), anti-double-stranded DNA (anti-dsDNA), and anti-Smith (anti-Sm). Statistical analyses included t-tests, Chi squared, Effect sizes, ANOVA, MANOVA, Euclidean distance measures, and logistic regression. RF-positive SS showed similar ACPA levels to RA in Welch’s T-tests, whereas RF-negative SS differed significantly in ANA and anti-dsDNA, consistent with a lupus-like profile within ANOVA. MANOVA and distance analyses indicated that RF-positive SS is immunologically closer to RA, while RF-negative SS forms a distinct cluster. Logistic regression supported moderate-low classification performance. Overall, these findings may support the hypothesis of an RF-associated SS subtype overlapping with RA and highlight the potential of RF testing to improve disease stratification and guide therapeutic decisions.

All data and code are publicly available (Zhuang, 2025).