Abstract
BACKGROUND: Existing literature has shown inconsistencies in the effects of exercise on hippocampal volume. Considering that including individuals without hippocampal atrophy may affect the significance of the results, this study systematically included older adults with preclinical hippocampal atrophy and patients diagnosed with neurodegenerative diseases associated with hippocampal atrophy. Using a meta-analysis approach, it aims to: quantify the effect of exercise intervention on hippocampal volume in high-risk older adults, and identify the moderating factors influencing intervention efficacy. METHODS: Databases including Embase, Ebsco, Cochrane Library, PubMed, and Web of Science were systematically searched. Two reviewers used the Cochrane Risk of Bias tool for quality assessment, and statistical analyses were performed using Manager 5.4 software. Sensitivity analyses were conducted via leave-one-out analysis and Stata 16 software to evaluate result robustness. Potential publication bias was assessed using funnel plot asymmetry and Begg test. RESULTS: Database searches identified 2066 studies, and 12 studies (774 participants) met the inclusion criteria after screening. The primary analysis showed no statistically significant effect of exercise intervention on hippocampal volume in older adults (SMD = 0.09, 95% CI: -0.05 to 0.22, P = .20). Subgroup analyses across the time span of exercise intervention, exercise modality, maximum target exercise intensity, health/disease grouping, left/right hippocampus grouping, and maximum duration of single-target exercise also showed no statistically significant effects (all P > .05). CONCLUSION: Current evidence does not support exercise intervention as an independent therapeutic modality for increasing hippocampal volume in older adults.