Background/Purpose: Despite effective control of inflammation in rheumatoid arthritis (RA), persistent and severe pain remains a clinical challenge. Dysregulated pain processing in the central nervous system (“central sensitization”) may underlie this disconnect. The PainDETECT questionnaire, originally developed to identify neuropathic pain, may capture features of central sensitization. We hypothesised that neuropathic-like pain at diagnosis predicts worse pain prognosis in early RA, independent of inflammation. We also explored associated brain activation patterns using functional neuroimaging.

Methods: In this prospective cohort study, adults with newly diagnosed RA in Oxfordshire completed the PainDETECT questionnaire at baseline. The primary outcome, bodily pain, was measured using the SF-36 Bodily Pain Subscale (BPS) at baseline, 3, 6, and 12 months. Inflammatory markers (CRP) were monitored longitudinally. Linear mixed-effects models examined associations between PainDETECT and BPS and CRP over time. A subset of 27 participants underwent task-based fMRI, assessing brain responses to evoked pressure pain.

Results: Among 158 participants (mean age 58.4 years; 35% male), higher baseline PainDETECT scores were associated with significantly worse bodily pain throughout follow-up (adjusted β= -0.43; 95%CI: -0.84 to -0.02; P=0.043), independent of socio-demographics, lifestyle, comorbidities, depression, and anxiety. No association was found between PainDETECT and CRP. On neuroimaging,higher PainDETECT scores correlated with increased activation in the left insula, dorsal anterior cingulate cortex (dACC), and left amygdala during evoked wrist pain.

Conclusion: Neuropathic-like pain at diagnosis predicts a worse pain prognosis in early RA, independently of inflammation. Early identification may support more targeted and effective pain management.

Schematic of the task-based functional MRI (tfMRI) protocol. Participants received alternating blocks of pressure pain (50 N) and light touch stimuli to the left wrist (typically affected in RA) and the left index finger nailbed (control site), during two 10-minute scan sessions on a 3T MRI scanner. Stimuli were applied manually using an MR-compatible algometer in a pseudorandomised order, guided by auditory cues. First-level contrasts modelled pain > touch at each site. Second-level contrasts compared wrist > nailbed activation, and group-level analyses examined associations with PainDETECT across the whole brain.

Lines represent individual raw patient trajectories of bodily pain (A) and C-reactive protein (CRP, B) over time, with mean values in bold. Shaded areas indicate 95% confidence intervals. Participants (N = 158, 91 scoring ≥13) were followed for one year. Neuropathic-like pain measured using the PainDETECT questionnaire (cutoff: ≥13 indicates probable neuropathic pain). Note higher values of SF36-BPS indicate more severe bodily pain. CRP values were taken from linked healthcare records during follow-up, and were grouped into average CRP in each time period.

Higher PainDETECT (PDQ) scores are associated with greater activity in the left Insula, dorsal anterior cingulate cortex (dACC), and left amygdala after whole brain correction. Whole brain group-level activation maps for painful vs. non-painful stimulation of the left wrist vs. left nailbed contrast with PainDETECT scores. Maps are overlaid on a standard MNI152 T1-weighted brain template and thresholded at Z > 3.1.

Higher PainDETECT scores at baseline are associated with significantly greater neuronal response in the left insula, dACC, and left amygdala. The y-axes show the induced effects of painful stimulation to the left wrist as a percentage BOLD signal change, defined as X = (Xpain (wrist) – Xno pain (wrist)) – (Xpain (nailbed) – Xno pain (nailbed)). Shaded areas represent the 95% confidence interval. Pearson correlation coefficients (r) are reported.

N&#3f27 participants who underwent neuroimaging.

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/neuropathic-like-pain-characteristics-predict-worse-pain-outcomes-in-early-rheumatoid-arthritis-a-prospective-cohort-study-with-embedded-neuroimaging-evaluation/