We read with interest the editorial by Soltys and colleagues about rheumatoid factors (RFs).1 They correctly stated that most naturally occurring RFs were of the IgM isotype while IgG RFs were thought to be associated with rhematoid arthritis (RA). It should be pointed out in this context that it is very difficult to measure IgG RF and this RF isotype can only be detected in approximately 50% of RA patients whereas IgA RF, usually in combination with IgM RF, can be detected in most patients with seropositive RA.2-4 It should also be noted that increased IgA RF is associated with severe manifestations in RA and this has been extensively reported by several different groups in the last 15 years.5-10

Furthermore, several reports have shown that RFs may be increased in serum many months or even years before clinical symptoms of …