Exercise for pain-related psychological outcomes in chronic musculoskeletal pain: a systematic review and meta-analysis

This review examined exercise effects on pain catastrophizing, kinesiophobia, fear-avoidance beliefs, and pain self-efficacy in chronic musculoskeletal pain. Pain intensity and strict disability were assessed only in trials reporting an eligible psychological outcome. This systematic review and meta-analysis followed PRISMA 2020 and was registered in PROSPERO on 7 May 2026 (CRD420261389453), before the formal database searches were conducted later that day. English-language randomized trials involving adults with chronic musculoskeletal pain were eligible. Random-effects meta-analyses used Hedges’ g, restricted maximum likelihood estimation, and Hartung–Knapp inference. Risk of bias was assessed using RoB 2 and certainty using GRADE. Forty-three randomized studies were included, and 20 independent studies contributed 49 outcome-specific effects to the primary analyses. Pain catastrophizing showed a small favorable average effect with low heterogeneity (6 studies; n = 332; Hedges’ g = 0.39, 95% CI 0.14 to 0.63; 95% prediction interval 0.07 to 0.70); certainty was low, and five of the six contributing studies involved participants with fibromyalgia. For pain intensity (15 studies; n = 1,017) and strict disability (12 studies; n = 836), certainty was very low and the 95% prediction intervals spanned effects in both directions (− 1.49 to 2.96 and − 0.88 to 1.84 , respectively), although the average estimates favored exercise (g = 0.74, 95% CI 0.14 to 1.34; and g = 0.48, 95% CI 0.02 to 0.94, respectively). Effects on kinesiophobia, pain self-efficacy, and fear-avoidance beliefs were less consistent. Certainty was very low for these three outcomes. The structured synthesis retained 92 additional non-primary study–outcome results; 34 favored exercise, 2 favored the comparator, and direction could not be determined reliably for 56. Corrected small-study-effect tests showed no statistical evidence of asymmetry, although publication bias could not be excluded. The effects of exercise varied across outcomes. Pain catastrophizing showed the most consistent psychological finding, although certainty was low and the evidence was concentrated in fibromyalgia. For pain intensity and strict disability, certainty was very low and prediction intervals spanned effects in both directions; although the average estimates favored exercise, effects across individual clinical settings remain uncertain. PROSPERO CRD420261389453.