Comment on: Clinical outcomes in psoriatic arthritis patients with rheumatoid factor positivity

Dear Editor, We read with considerable interest the recent article by Mehta et al. evaluating clinical impact of rheumatoid factor (RF) positivity in psoriatic arthritis (PsA) [1]. While the authors present an impressive longitudinal dataset, we are concerned that the analytical framework does not support the causal inferences implied, and several acknowledged limitations are not fully interrogated in terms of their methodological consequences, thereby weakening the causal interpretation of the findings. First, the exposure definition, ever RF positivity, remains a central concern, and introduces a substantial risk of misclassification bias. Although the authors appropriately acknowledge that most RF-positive cases were transient and low-titer, the analytical implications of this are likely underestimated. Low-level RF positivity is known to occur in healthy individuals, other connective tissue diseases, and also increases with age and smoking exposure, often without clear pathological relevance [2]. Pooling transient, low-level RF positivity, together with persistent seropositivity introduces non-differential misclassification that is likely time-dependent. In longitudinal settings, such misclassification does not merely attenuate associations but can produce biased and unstable estimates, particularly when exposure status is updated over time without explicit temporal structuring or lagged definitions [3]. Consequently, the reported associations may partly reflect artefactual signals arising from exposure instability rather than a biologically meaningful effect.