Comparison of the Efficacy of Different Traditional Chinese Exercises in the Treatment of Chronic Non-Specific Low Back Pain: A Bayesian Network Meta-Analysis

Background: Chronic non-specific low back pain (CNSLBP) is a disabling condition worldwide, with exercise therapy as its first-line intervention. Traditional Chinese exercises (eg, Tai Ji, Baduanjin, Wuqinxi) have gained attention, but their comparative efficacy remains unclear. Objective: To compare the efficacy of six traditional Chinese exercises—Tai Ji, Yijinjing, Wuqinxi, Baduanjin, Chan-Chuang Qigong, and Liuzijue—in patients with CNSLBP via Bayesian network meta-analysis. Methods: We searched Chinese (CNKI, Wanfang, VIP, SionMed) and English (PubMed, Web of Science, Cochrane Library, Embase) databases up to December 3, 2025. A total of 24 randomized controlled trials (RCTs) with 2,067 participants were included. The overall methodological quality was moderate, with common limitations in blinding and allocation concealment. A Bayesian random-effects model was used to compare interventions based on visual analogue scale (VAS) and Oswestry disability index (ODI) scores, with ranking performed via the surface under the cumulative ranking curve (SUCRA). Conventional therapy (CT) was defined as rehabilitation training (eg, core muscle strengthening, suspension training, rehabilitation gymnastics), Western medicine, manipulation, or acupuncture. Results: For pain reduction (VAS), Wuqinxi combined with CT (SUCRA=0.80), Chan-Chuang Qigong combined with CT (0.77), and Yijinjing alone (0.69) ranked highest. In pairwise comparisons, these interventions showed statistically significant improvements over conventional therapy alone (P < 0.05). For functional improvement (ODI), Liuzijue combined with CT (0.99), Baduanjin combined with CT (0.76), and Tai Ji combined with CT (0.72) performed best, with significant differences versus CT (P < 0.05). Notably, combined therapies generally outperformed single-modality exercises. Meta-regression suggested that longer intervention durations enhanced the effects of Baduanjin and Wuqinxi. Conclusion: Traditional Chinese exercises, particularly Wuqinxi and Liuzijue when combined with conventional therapy, show promising effects for pain relief and functional improvement in CNSLBP. However, the evidence quality is moderate to low, limited by heterogeneity and methodological constraints. Future studies should prioritize higher-quality designs, longer follow-up, and standardized intervention protocols. Keywords: low back pain, exercise therapy, mind-body therapies, network meta-analysis, Bayes theorem, meta-analysis as topic, treatment outcome, rehabilitation