Abstract Background/Aims Pain is often persistent in patients with inflammatory arthritis (IA). Deciphering whether this pain is primarily inflammatory or non-inflammatory (via predominantly peripheral or central nervous system mediated mechanisms) offers opportunities for mechanism-based care. This study prospectively investigated the relative contribution of peripherally versus centrally mediated mechanisms of pain from newly diagnosed IA over two years. Methods Patients with newly diagnosed IA and baseline pain scores ≥3 were followed across 4 timepoints over two years. Assessments included: disease activity (Disease Activity Score-28, musculoskeletal ultrasound); quality of life (Musculoskeletal Health questionnaire (MSK-HQ); mental health status (Patient Health Questionnaire Anxiety, Depression Scale), and pain characteristics (fibromyalgia criteria, painDETECT, Static and Dynamic Quantitative Sensory Testing, QST). Mixed-effects regression models examined longitudinal associations. Results Among 66 participants (all baseline pain NRS ≥3), pain decreased significantly during the first six months, with smaller, non-significant reductions thereafter; at 24 months, 49% still reported pain (NRS ≥3). Centrally mediated pain markers improved (fibromyalgia severity -2.7; PainDETECT -2.7; PHQ-15 -1.9; GAD-7 -2.3; PHQ-9 -1.7), as did inflammation (CRP -4.9; SJC -1.7; EGA -12.6) and MSK-HQ (+10.3). Joint PPT increased, but trapezius PPT, TSP and CPM did not significantly change. In mixed-effects models over two years, controlling for prior pain, inflammation was not associated with subsequent pain (AME -0.05, 95% CI -0.67 to 0.56), whereas higher centrally mediated pain markers predicted less pain reduction (AME 1.41, 95% CI 0.83-1.99, p 0.001). Participants with persistent pain had higher baseline and longitudinal scores for centrally mediated pain (fibromyalgia severity, painDETECT, mental health, fatigue), despite similar baseline inflammation and comparable reductions over time. Baseline centrally mediated pain, but not inflammation, predicted persistent pain (AME 1.32, 95% CI 0.83-1.82, p 0.001); baseline fatigue, painDETECT, fibromyalgia severity, PHQ-15, PHQ-9, and GAD-7 were the strongest predictors (Table 1). Conclusion In our IA cohort followed from diagnosis, while inflammation reduced comparably across patients, centrally mediated mechanisms primarily contributed to pain persistence. Early identification and treatment of such mechanisms may be key to optimising long-term pain outcomes. Disclosure Z. Rutter-Locher: Grants/research support; NIHR301674. S. Norton: None. B. Menon: None. L. Taams: None. K. Bannister: None. B. Kirkham: None.
Rutter-Locher et al. (Wed,) studied this question.