Physiological Adaptations and Serum-Based Biomarker Dynamics During Multimodal Rehabilitation in Chronic Pain: Analysis of a Prospective Cohort Study

Background: Chronic pain is a multifactorial condition for which interdisciplinary multimodal rehabilitation is guideline-recommended, yet the biological mechanisms underlying treatment response remain incompletely understood and validated predictive biomarkers have not been established. Objective: This exploratory prospective cohort study examined clinical outcomes and circulating biomarker changes, encompassing the endocannabinoid system (ECS), inflammatory mediators, stress-regulatory markers, and metabolic parameters, in 410 patients with chronic pain of predominantly musculoskeletal etiology undergoing a standardized five-week rehabilitation program. Materials and Methods: Pain intensity and affective pain were assessed at baseline and end of rehabilitation; global performance of treatment (GPT) was additionally recorded. Serum analyses included anandamide (AEA), 2-arachidonoylglycerol (2-AG), IL-6, cortisol, IGF-1, BDNF, and leptin. Biomarker-outcome associations were examined via multiple regression analyses adjusted for demographics and biological and clinical confounders. Results: Statistically significant reductions were observed in pain intensity (−0.785 points, NRS; p < 0.001) and affective pain (−0.750 points; p < 0.001). IL-6 was associated with pain outcomes across time points. Higher baseline 2-AG independently predicted lower end-of-rehabilitation pain intensity, affective pain, and more favorable GPT. Greater AEA reductions were associated with favorable GPT. Conclusions: Baseline 2-AG emerges as a candidate predictor of treatment response, with lower pre-treatment levels potentially reflecting reduced stress-adaptive capacity, supporting inclusion of ECS markers in future controlled biomarker studies.