Objectives. Smoking is associated with false-positive rheumatoid factor (RF). We explored the dose-response relationship of this association, using smoking history and serum cotinine and thiocyanate concentrations as measures of tobacco exposure. Methods. A total of 6947 men and women aged 30 yr or over and free of clinical arthritis were included in the Mini-Finland Health Survey carried out between 1978 and 1980. Detailed histories of smoking and RF (sensitized sheep cell agglutination test)

more »

... ere obtained in the basic examination. In 2000, serum cotinine and thiocyanate were determined from serum samples collected at baseline and stored at À20 C. A cut-off point of 100 lg/l was used for serum cotinine and 10 lmol/l for thiocyanate to indicate active smoking. Results. There was a close association between smoking and strongly positive RF. After adjustment for age, sex, coffee consumption and region, the odds ratios (95% confidence intervals in brackets) in current smokers and in those who had quit smoking were 3.94 (2.04-7.61) and 2.71 (1.33-5.53), respectively, compared with those who had never smoked. Among current smokers, the intensity, duration or tertiles of pack-years of smoking were not related to RF. No relationship between serum cotinine or thiocyanate and RF positivity was observed within the subgroups of current smokers and those who had quit. Among those who reported that they had never smoked but who nevertheless had serum cotinine levels at least 100 lg/l, the adjusted odds ratio of strongly positive RF was 4.48 (1.48-13.50) compared with people who had never smoked and whose serum cotinine levels were less than 100 lg/l. Conclusions. The results are not in line with the hypothesis of a dose-response relationship between smoking exposure and RF positivity. all rights reserved Rheumatology Key messages There is a close association between smoking and RF in non-rheumatoid subjects. However, no dose-response relationship seems to exist between smoking exposure and RF positivity.