Prediction of articular destruction in rheumatoid arthritis: disease activity markers revisited

To assess the predictive value for joint damage progression of commonly used disease activity or process measures in rheumatoid arthritis (RA).

Seventy-two patients fulfilling the American Rheumatism Association criteria for RA were assessed twice yearly for 2 years. Primary outcome variables were progression of articular destruction, evaluated by Sharp's method, for 6, 12, 18, and 24 month periods.

Regression analysis, using random effects linear models, showed that only C-reactive protein, alpha 1-acid glycoprotein, iron, and erythrocyte sedimentation rate were significantly, but not independently, associated with 6 month radiographic progression. Traditional clinical measures were not predictive. No assessed marker was able to predict longer term outcome (12 or 18 month joint damage progression). Recent onset disease and older age were also associated with more severe radiographic progression.

The lack of association between clinical measures and laboratory markers as predictors of the progression of articular destruction is further evidence of the need to reconsider processes and outcomes in RA. This study also suggests that clinical measures and laboratory markers probably do not reflect the same underlying process, arguing against gathering these measures under the same heading of "disease activity measures".