LAUSR

OBJECTIVES This study investigates longitudinal patterns, predictors and long-term impact of pain in axial spondyloarthritis (axSpA), using clinical and self-tracking data. METHODS The presence of multisite pain (MSP), affecting at least 6/9 body regions using a Margolis pain drawing, and subsequent chronic widespread pain (CWP), MSP at more than one timepoint, was assessed in a cohort of axSpA patients. Incident MSP (MSP at two consecutive visits or more), intermittent MSP (MSP at two or more non-consecutive visits) and persistent MSP (MSP at each visit) were described. Demographic, clinical and self-tracking measures were compared for the CWP versus non-CWP groups using Students t-test, Wilcoxon-Mann-Whitney and χ2 test for normal, non-normal and categorical data respectively. Predictors of CWP were evaluated using logistic regression modelling. RESULTS 136 patients, mean clinical study duration of 120 weeks (range 27-277 weeks) were included, with sufficient self-tracking data in 97 patients. 68 (50%) patients reported MSP during at least one clinical visit: 8 (6%) incident MSP; 16 (12%) persistent MSP, and 44 (32%) intermittent MSP. 46 (34%) of the cohort had CWP. All baseline measures of disease activity, function, quality of life, sleep disturbance, fatigue and overall activity impairment were significant predictors of the development of CWP. BASDAI and BASFI scores were significantly higher in those with CWP and self-tracking data revealed significantly worse pain, fatigue, sleep quality and stress. CONCLUSIONS The development of CWP is predicted by higher levels of disease activity and burden at baseline. It also impacts future disease activity and wellbeing.