The pain that inflammation does not explain in SpA and PsA Save

Clinicians often come across the scenario of a patient with well-controlled spondyloarthritis where the ASDAS in the low disease activity range, CRP normalised, joints clinically quiet. However, the patient returns to the clinic still reporting significant pain. Their rheumatologist adjusts the biological, waits, returns. The pain persists. This scenario reflects a dimension of these diseases that standard disease activity scores were not designed to measure: the contribution of neuropathic-like and nociplastic pain.

At EULAR 2026, the NEUPIA study (OP0317), a prospective observational investigation conducted at KU Leuven and University Hospitals Leuven, provides the most detailed longitudinal characterisation to date of these non-inflammatory pain phenotypes in Psoriatic Arthritis (PsA), Spondyloarthritis (SpA), and rheumatoid arthritis (RA), and their associations with anxiety, depression, and sleep disturbance over six months.

Neuropathic-like pain refers to pain with sensory characteristics typical of peripheral or central nerve sensitisation with burning, shooting, pins and needles. This is assessed using the PainDETECT questionnaire (score ≥19). Nociplastic pain represents a third mechanistic category beyond nociceptive (tissue damage) and neuropathic (nerve damage) pain. Pain arising from altered nociception without clear evidence of tissue damage or nerve injury is assessed using the Widespread Pain Index (WPI) criteria. Both are increasingly recognised as major contributors to symptom burden in inflammatory arthritis but data on their longitudinal behaviour and their relationship to mental health and sleep across specific diagnoses have been limited.

NEUPIA enrolled 414 patients across three diagnoses: psoriatic arthritis (PsA, 40.6%), spondyloarthritis (SpA, 31.6%), and rheumatoid arthritis (RA, 27.8%). At baseline and six months, investigators assessed PainDETECT scores, WPI criteria, Hospital Anxiety and Depression Scale (HADS), and Pittsburgh Sleep Quality Index (PSQI). The study was designed to capture not just prevalence but the longitudinal trajectory of each pain phenotype and its relationship to psychological and sleep outcomes.

Across every measure, SpA patients demonstrated the highest pain-related psychological and sleep burden. Nociplastic pain was significantly more prevalent in SpA (16.0%) than in PsA or RA (p = 0.013), and poor sleep quality was markedly more common in SpA patients with neuropathic-like pain (41.7%) compared with PsA patients with the same pain phenotype (27.1%). At six months, SpA patients showed a statistically significant rise in depression scores among those with nociplastic pain, reaching 70.6%. 

One of the most important findings from NEUPIA is that when patients changed their pain phenotype classification over six months (shifting from neuropathic-like or nociplastic to normal, or vice versa), the majority showed stable HADS and PSQI scores. In other words, the pain mechanism itself appears to change independently of anxiety, depression, and sleep quality.  

The NEUPIA data make an argument for pain phenotyping in inflammatory arthritis clinics particularly in SpA and PsA, where the gap between objective disease control and patient-reported pain is frequently encountered. Tools like PainDETECT and WPI criteria are not yet standard components of rheumatology follow-up, but their application could identify patients whose persistent pain reflects central sensitisation rather than ongoing synovial inflammation and who therefore need a fundamentally different therapeutic approach.