To study the characteristics of temporal summation (TS) and conditioned pain modulation (CPM) in fibromyalgia (FM) patients, we systematically searched Pubmed and EMBASE for studies using TS or CPM comparing FM patients with healthy controls. We computed Hedges' g, risk of bias, sensitivity analysis, and meta-regression tests with 10,000 Monte-Carlo permutations. Twenty-three studies (625 female and 23 male FM patients and 591 female and 81 male healthy controls) were included. The meta-analyses showed an effect size of .53 for TS (P < .001; 95% confidence interval = .23-.83), which is a 68% relative difference between patients and controls, and of .57 for CPM (P < .001; 95% confidence interval = -.88 to -.26), representing a 65% relative difference between the groups. The qualitative analyses revealed large heterogeneity between study protocols. Although studies were of low risk of bias, lack of blinding was substantial. Sensitivity analysis and meta-regression identified type and site of stimulation, age, lab, sample size, and medication control as important sources of between-study variability. We showed a significant alteration of pain modulation mechanisms in FM patients.
Perspective: This novel meta-analysis provides evidence for defective endogenous pain modulation in FM patients. We explored the effect of covariates on between-study variability in these paradigms. These biomarkers may aid in diagnosis, and treatment of patients. However, validation requires further investigation under strict methodological settings, and into individual patient covariates.
Keywords: Conditioned pain modulation; chronic pain; diffuse noxious inhibitory control; fibromyalgia; temporal summation.
Copyright © 2018 The American Pain Society. Published by Elsevier Inc. All rights reserved.