Efficacy of non-pharmacological interventions on central sensitisation to pain: A systematic review and network...

Central sensitisation is a key mechanism underlying many chronic pain conditions. Although non-pharmacological interventions such as exercise, manual therapy, and pain neuroscience education have been recommended by some experts, the comparative effectiveness of these therapies on central sensitization-specific outcomes remains unclear.

To compare and rank the efficacy of various non-pharmacological interventions on validated indices of central sensitisation in adults with chronic pain.

A systematic review and Bayesian network meta-analysis of randomized controlled trials was conducted. Databases were searched from inception to September 2025. Interventions included manual therapy, pain neuroscience education, exercise, virtual reality, electrical nerve stimulation, and their combinations. Primary outcomes were the patient-reported Central Sensitisation Inventory as a measure of subjective symptom burden, and pressure pain threshold as an objective quantitative sensory test. Standardized mean differences (SMD) with 95% credible intervals and surface under the cumulative ranking curve values were calculated. Inconsistency, heterogeneity, meta-regression, sensitivity, and publication bias analyses were performed. Confidence in evidence was evaluated using the Confidence in Network Meta-Analysis framework.

Forty-two trials (2811 participants) were included, with a median intervention duration of 6 weeks. For the Central Sensitisation Inventory, pain neuroscience education showed the largest effect (SMD = -1.5; 95% CrI: -3.2 to 0.27), though the credible interval included the null. For pressure pain threshold, pain neuroscience education demonstrated a large, statistically significant effect (SMD = 4.0; 95% CrI: 2.7 to 5.3; high certainty). No statistically significant inconsistency or publication bias was detected. Meta-regression identified no effect modification by pain type or baseline severity. Confidence in Network Meta-Analysis indicated high-certainty evidence for pain neuroscience education on pressure pain threshold, but low-to-very-low certainty for most Central Sensitisation Inventory comparisons.

Pain neuroscience education is effective for both subjective and objective central sensitisation indices, though evidence for patient-reported outcomes is low certainty. Exercise effects are outcome-dependent, revealing a "therapeutic dissociation" between physiological adaptation and symptom report. Clinical management should be goal-oriented, integrating pain neuroscience education with other modalities based on patient profiles. Further high-quality research is needed.

This review was prospectively registered with PROSPERO (CRD420251266071) and received no dedicated funding.