Background Chronic low back pain (CLBP) is often accompanied by anxiety and depression, which hinders clinical treatment. Therefore, this study conducts a Bayesian network meta-analysis to compare the effectiveness of different non-drug interventions in relieving pain and related emotional symptoms in CLBP patients. Methods This review was prospectively registered in PROSPERO (CRD420251066414). As of June 2025, PubMed, Embase, Cochrane Library, Web of Science, CINAHL, Scopus, CNKI, and Wanfang databases have been searched to collect randomized controlled trials concerning non-pharmacological interventions for CLBP. The main outcomes are pain, anxiety, and depression. The scoring directions of all outcome scales were consistent. To account for differences in measurement tools, effect sizes are expressed as standardized mean differences (SMD), estimated using a Bayesian random-effects model based on change scores, and reported with 95% credible intervals (CrIs), adjusted using Hedges' g. Comprehensive sensitivity analysis, subgroup analysis, and regression analysis were carried out to explore the source of heterogeneity and the factors affecting the treatment results. Results Twenty-four trials were ultimately included (n = 1828), involving 9 non-pharmacological interventions. The analysis of the results shows that mind-body exercises (MBE) demonstrated favorable effects for pain relief SMD: −1. 55, 95% CrI (−2. 50, −0. 59) and depression and anxiety reduction SMD: −1. 14, 95% CrI (−1. 74, −0. 54), SMD: −1. 38, 95% CrI (−1. 95, −0. 81). Structured exercise (SE) was also associated with relief in pain SMD: −1. 02, 95% CrI (−1. 94, −0. 09), depression, and anxiety SMD: −0. 77, 95% CrI (−1. 35, −0. 19), SMD: −0. 69, 95% CrI (−1. 25, −0. 14). Integrated rehabilitation therapy (IR) showed a potential benefit for improving depression SMD: −0. 74, 95% CrI (−1. 28, −0. 18). Considerable heterogeneity was observed across all outcome measures (pain: 86. 24%; depression: 70. 64%; anxiety: 66. 08%). Additional subgroup interaction analyses did not identify any factors influencing treatment outcomes. Meta-regression suggested that the number of interventions may influence pain, whilst the type of intervention may influence anxiety. Conclusion MBE and SE may help alleviate pain, anxiety, and depression among CLBP individuals, whilst IR may have a modest beneficial impact on depression. Given the high heterogeneity across different outcome measures, these conclusions should be evaluated with caution, and further high-quality research is required to substantiate them. Systematic Review Registration https: //www. crd. york. ac. uk/prospero/displayᵣecord. php? ID=CRD420251066414, PROSPERO, CRD420251066414.
Chen et al. (2026) studied this question.