Predictors of Change in Bodily Pain in Early Rheumatoid Arthritis: An Inception Cohort Study

OBJECTIVE To investigate possible predictors for lack of pain improvement after 1 year of treatment for early rheumatoid arthritis (RA). METHODS The Early Rheumatoid Arthritis Network (ERAN) database was used for analysis of baseline and 1-year pain data. The ERAN is a hospital-based inception cohort of 1,189 people. Short Form 36 questionnaire bodily pain scores were used to calculate change in pain at 1 year as the outcome. The proportion of the Disease Activity Score in 28 joints (DAS28) attributable to patient-reported components (joint tenderness and visual analog scale score; DAS28-P) at baseline was derived as a predictor. Predictors of less improvement in pain were investigated using adjusted odds ratios (OR(adj) ) generated by logistic regression, adjusting for 14 additional clinical and demographic covariates. RESULTS Greater pain at baseline was associated with sex, high DAS28, worse mental health, and smoking. Most patients with early RA reported incomplete improvement in bodily pain after 1 year. The DAS28-P index did not significantly change in the patients whose disease remained active. Less improvement in pain was predicted by female sex (OR(adj) 3.41, 95% confidence interval [95% CI] 1.35-8.64) and a high DAS28-P index at baseline (OR(adj) for tertiles 2.09, 95% CI 1.24-3.55). Other conventional RA risk factors did not predict pain changes. CONCLUSION The factors most likely to predict less improvement in pain in early RA are female sex and a high DAS28-P index. A high DAS28-P index may reflect greater contributions of noninflammatory factors, such as central sensitization, to pain. Strategies in addition to inflammatory disease suppression may be required to adequately treat pain.