Baseline symptom severity profiles and response to physiotherapy treatment in chronic musculoskeletal pain: A cluster analysis in primary care

Background Chronic musculoskeletal pain (CMP) is a heterogeneous condition in which overall treatment effects may obscure differences across patient subgroups. Objective To identify distinct baseline symptom profiles and examine their association with baseline-adjusted post-treatment outcomes, responder status, exploratory multidomain negative change, and response trajectorie s following a structured active pain coping program in primary care. Methods This was a prospective observational cohort study ( n = 169). Participants were enrolled in a 12-week physical therapy program combining pain neuroscience education (PNE) and therapeutic exercise. Two-step cluster analysis identified baseline profiles from pain catastrophizing (PCS), kinesiophobia (TSK-11), disability (RMDQ), and central sensitization (CSI). Pre–post changes, ANCOVA-adjusted endpoints, responder rates, and K-means response trajectories were examined. Results Two profiles emerged: Moderate (57.4%) and Severe (42.6%). Participants in both profiles improved significantly, with larger absolute reductions in the Severe profile (PCS responders: 77.6% vs 47.8%). After adjusting for baseline values, between-group differences were non-significant ( p ≥ 0.05) across all domains, with similar adjusted endpoint levels in both profiles. The exploratory multidomain negative-change criterion was met by 21.6% of participants and was concentrated in the Moderate profile (88%; OR = 3.99, 95% CI: 1.36–11.70). Response-pattern analysis identified three trajectories: Global Responders (broad improvement across domains; 7.1%), Limited Responders (modest symptom change with worsening mental health; 45.9%), and Moderate Responders (intermediate improvements; 46.9%); the distribution of these trajectories was significantly associated with baseline profile. Only 31.0% of participants met the prespecified ideal responder criterion (simultaneous improvement in symptoms and quality of life). Conclusion Both profiles converged toward comparable functional endpoints, making baseline severity a poor gatekeeping criterion for access to this care. Nonetheless, exploratory multidomain negative change was more frequent in the Moderate profile, and concurrent symptom and quality-of-life recovery was uncommon, underscoring the need for systematic mid-treatment monitoring and outcome definitions that extend beyond symptom scores.