A Bayesian network meta-analysis of specific exercise modalities for depressive symptoms in mild cognitive impairment

BackgroundPatients with mild cognitive impairment (MCI) frequently experience depressive symptoms, which are linked to accelerated cognitive decline, poorer quality of life, and an increased risk of progression to Alzheimer's disease. Exercise is considered a promising non-pharmacological strategy for mood improvement and brain health; however, the specific effects of different exercise modalities remain insufficiently defined.ObjectiveTo compare the relative efficacy of specific exercise interventions on depressive symptoms in patients with MCI.MethodsRandomized controlled trials (RCTs) evaluating the effects of exercise interventions on depressive symptoms in patients with MCI were systematically searched. Standardized mean differences (SMD) were calculated to pool outcomes across different depression scales. A Bayesian network meta-analysis was conducted to integrate direct and indirect evidence, and interventions were ranked using the surface under the cumulative ranking curve (SUCRA). Network consistency and model fit were also assessed.ResultsA total of 19 RCTs involving nine exercise modalities were included. Compared with the control group, dance exercise demonstrated a statistically significant improvement in depressive symptoms (SMD = -0.51, 95% credible intervals: -0.84 to -0.16), representing a moderate effect size, and showed a relatively high-ranking probability (SUCRA = 80.71%). Overall heterogeneity was moderate, and local inconsistency was observed in some comparisons.ConclusionsExercise interventions may contribute to the alleviation of depressive symptoms in patients with MCI, with dance exercise showing relatively greater therapeutic potential. However, these findings should be interpreted with caution in light of evidence inconsistency, and further high-quality studies are warranted to substantiate these results.