Background A number of studies have shown that a high level of disease activity is associated with more rapid radiological progression. In those studies time-averaged or time-integrated estimates for disease activity have been used. The concept of time-averaged disease activity does not entirely reflect reality, because disease activity is rarely constant over time. An inherent shortcoming of time-averaged estimates is that they do not really predict radiological outcome.
Objectives To investigate, using multivariate longitudinal regression analysis, whether disease activity can predict radiological progression with a fixed time lag of 6 months.
Methods 5 years follow-up data of RA-patients having participated in the COBRA trial (a 56-weeks double blind controlled trial in 146 patients with early, DMARD-naive, RA which compared COBRA-combination therapy with sulfasalazine monotherapy) were used. After the completion of the trial, patients had been followed for another 4 years, and X-rays of hands and feet, as well as assessments for disease activity, were obtained with 6 months intervals. X-rays were scored according to the van der Heijde-modified Sharp-method in a chronological order. The 28 joints disease activity score (DAS28) was calculated. Generalised estimating equations (GEE) was used to study the relationship between DAS28 and time (explanatory variables) and radiological damage (dependent variable). Every individual damage score was adjusted for its previous score (6 months earlier) (autoregression) in order to study radiological
Results The interpretation of the beta’s in the Table 1 is that each additional DAS28 point, measured at any time point, predicts a mean radiological progression of 0.52 Sharp-points during the following 6 months interval. The negative sign for time indicates that radiological progression in this study decreased over time.
Conclusion By using GEE we have shown that DAS measured at any time point during a 5 years follow-up predicts radiological progression during the following 6 months. GEE is an appropriate method to study longitudinal relationships between outcome- and explanatory variables in RA.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.