We read with interest the report by Turesson et al on the incidence of vasculitis and other extraarticular manifestations in rheumatoid arthritis (RA) (1). In the long-term study of a community-based cohort of patients with RA, no evidence was found for a decrease in the incidence of severe vasculitis.
As the authors pointed out, their results differ from those reported by Watts et al, who observed a >3-fold decreased incidence of systemic rheumatoid vasculitis from 1988 to 2002 in Norwich, UK (2). Another study, which may have not been mentioned in the report by Turesson et al due to space limitations, evaluated the incidence of severe extraarticular rheumatoid manifestations by reporting the rates of hospitalization in the US (according to the National Institute of Arthritis and Musculoskeletal and Skin Diseases/National Institutes of Health) (3). Again, the rate of hospitalization due to systemic vasculitis decreased by one-third from 1983 to 2001.
Turesson et al give some potential explanation for the discrepancies with the previous results. We would like to suggest another possibility. Although it was extensive, the analysis performed by Turesson et al was restricted to the first decade of illness. All reported studies of systemic rheumatoid vasculitis have highlighted the fact that vasculitis usually occurs several years after the onset of arthritis. An analysis of 398 cases of systemic rheumatoid vasculitis reported in the literature through 1990 revealed a mean duration of RA of 11 years before vasculitis occurred (4). For example, in a hospital-based study of 32 patients with systemic rheumatoid vasculitis with peripheral neuropathy, the mean duration of RA was 16 years (range 2–47 years) (5). Had we restricted the study to only those cases observed within the first decade of RA, more than half (56%) of these complications would have been missed.
Before drawing any conclusion from the study conducted by Turesson et al, we would like to ask the authors to provide the incidence of rheumatoid vasculitis in their cohort of patients without limiting the analysis to the first decade of illness. With the complete data, which will increase the number of cases, especially in patients for whom the diagnosis had been made several decades ago, the incidence might no longer remain as stable as initially reported.