Résumé
In a cohort of early rheumatoid arthritis (RA) patients, we aimed to determine and characterise fatigue trajectories over 10 years of follow-up and identify predictors of trajectory membership.Methods: We selected patients fulfilling the 2010 ACR/EULAR criteria for RA included in the ESPOIR cohort. We used a cluster analysis to obtain fatigue (assessed by fatigue visual analogue scale) trajectories over the course of 10 years from enrolment. Chi-square tests or ANOVA were performed to evaluate differences of baseline variables between fatigue trajectories. Using a multinomial logistic regression we were able to identify predictors of trajectory membership.Results: We analysed 598 patients with mean disease duration at enrolment of 26.2±40.9 days. Cluster analysis revealed 3 trajectories: high (18%), moderate (52%) and low fatigue (30%). Compared to patients with moderate or low fatigue trajectory, patients with high fatigue trajectory were predominantly women and reported significantly higher duration and intensity of morning stiffness, HAQ score, tender joints count, levels of pain, number of awakenings due to arthritis, frequency of fibromyalgic RA, levels of physician and patient global assessment, more frequent sleep problems, and increased psychological distress. Female patients with pain, psychological distress and presence of sicca symptoms had a higher risk of being in the high trajectory group.Conclusions: These findings suggest that levels of fatigue are rather stable over time in each trajectory. Baseline clinical measures and baseline patient-reported measures of functional status better distinguished the three fatigue trajectories. We did not find any differences between trajectories in baseline laboratory measures. Inflammatory activity was not a predictor of being in the high trajectory fatigue group.