Comparative effects of structured exercise protocols on depression and anxiety symptoms: a network meta-analysis.

内容摘要

To systematically compare the intervention effects of eight common structured exercise modalities on depressive and anxiety symptoms in adults using a network meta-analysis approach, providing evidence-based insights for developing precision exercise prescriptions in mental health. A computer search was conducted across PubMed, Web of Science, Embase, and the Cochrane Library databases, covering the period from the database establishment to March 31,2026. Randomized controlled trials comparing different exercise modalities for treating adult depressive or anxiety symptoms were included. Literature quality was assessed using the Cochrane Collaboration's Bias Risk Assessment Tool (Revised Version), with statistical analyses performed using Stata 19.0 and RevMan 5.4. The standardized mean difference (SMD) was used as the effect measure, consistency was evaluated by the node splitting method, and the cumulative rank-sum area under the curve (SUCRA) was calculated to rank intervention efficacy. Publication bias was assessed and corrected using the trimming method and meta-regression. A total of 22 studies involving 23 randomized controlled trials with 1,830 participants were included, encompassing eight exercise modalities: yoga, Tai Chi, Pilates, resistance training, aerobic exercise, combined exercise, high-intensity interval training, and moderate-intensity continuous training. Traditional meta-analysis demonstrated that exercise intervention groups showed significantly better improvements in both depressive symptoms (SMD = -0.67,95% CI: -0.97 to-0.37) and anxiety symptoms (SMD = -0.77,95% CI: -1.12 to-0.41) compared to the control groups (both P <0.001). The network meta-analysis results demonstrated that all exercise modalities were effective. In the SUCRA probability ranking, yoga ranked first in both depression (SUCRA&#xa0;=&#xa0;68.8%) and anxiety (SUCRA&#xa0;=&#xa0;72.2%) improvement. However, the differences in effect sizes between yoga and moderate-intensity continuous training (MICT) or high-intensity interval training (HIIT) were not statistically significant (all 95% confidence intervals included 0), indicating no clear superiority among the three interventions; thus, this ranking should be regarded as exploratory rather than confirmatory evidence. The top three interventions for depression symptom improvement were yoga, MICT (56.7%), and HIIT (54.6%), while the top three for anxiety symptom improvement were yoga, HIIT (57.4%), and MICT (56.2%). Subgroup analyses revealed no statistically significant moderating effects of intervention duration or age (all P>0.05), although the effect sizes were larger in the elderly group (&#x2265;60 years) compared to other age groups. Consistency testing indicated a reliable evidence network (P>0.05). Egger's test suggested potential publication bias (depression P&#xa0;=&#xa0;0.017, anxiety P&#xa0;=&#xa0;0.010), but the meta-analysis did not incorporate missing studies, and meta-regression did not detect small-sample effects; the direction and significance of the effects remained unchanged, rendering the conclusions robust. Different exercise modalities exhibit beneficial effects on both depression and anxiety symptoms. Yoga, MICT, and HIIT all demonstrated significant potential in alleviating both symptoms; however, the differences in efficacy among the three interventions were not statistically significant, making them all viable prioritized exercise modalities in clinical practice. Clinical selection should be based on a comprehensive evaluation of the patient's dominant symptom cluster, tolerance, preferences, and safety profile to develop individualized exercise regimens. Since only one study included Pilates, its independent effect remains to be validated. https://www.crd.york.ac.uk/prospero/, identifier.