OBJECTIVE
To carry out a cross sectional case-control study of the risk factors for articular destruction in a large sample of patients with a long history of rheumatoid arthritis (RA), presupposing that the variables we measured were unrelated to the duration of disease.
METHODS
Each inpatient with RA admitted to our department from January 1, 1985, to December 31, 1995, underwent standard examination, laboratory tests, and hand roentgenograms. We carried out a radiographic cross sectional study on 287 of them. Radiographic evaluation was performed by the same observer (correlation coefficient 0.97) using the modified Sharp method. To investigate an association between articular destruction and prognostic variables, a matched analysis of the case-control data and calculation of the odds ratio (OR) with 95% confidence intervals (CI) were carried out. For each patient with severe articular destruction, patients hospitalized during the study period with the same disease duration but without severe articular destruction were included as controls. The sample size was chosen to show an OR > 2 (1-alpha = 95%; 1-beta = 80%).
RESULTS
The risk of articular joint destruction was higher in women than in men (OR 2.72, CI 1.17-7.9, p<0.023), whereas age at onset or the presence of HLA-DR4, antiperinuclear factor, or antikeratin antibodies was not sufficiently strongly associated with the process of articular destruction to be considered relevant prognostic markers.
CONCLUSION
We conclude that female sex is significantly associated with a higher risk of articular destruction.