神経変性疾患における運動誘発性脳由来神経栄養因子の変化：ベイジアンネットワークメタアナリシス

ObjectivesThis study aimed to compare the effects of different exercise interventions on brain-derived neurotrophic factor (BDNF) levels in patients with neurodegenerative diseases and to explore regulatory factors.MethodsSearched PubMed, Scopus, Web of Science Core Collection, CNKI and Cochrane Library databases up to March 15, 2025. Bayesian network meta-analysis was conducted using R software, and meta-regression analyzed the moderating effects of training period and frequency.Results42 randomized controlled trials covering 1482 patients were included. The Surface Under the Cumulative Ranking (SUCRA) indicated that stretching training (SUCRA = 78.92) and high-intensity interval training (SUCRA = 69.73) were ranked higher than other exercise modalities and exhibited more favorable effect on BDNF enhancement, although neither demonstrated statistically significant superiority over the blank control. In contrast, combined training (SUCRA = 35.58), aerobic training (SUCRA = 35.17), and resistance training (SUCRA = 12.98) showed relatively lower potential for BDNF enhancement (blank control SUCRA = 67.62). Meta-regression analysis showed that the effect of combined training was significantly and positively correlated with intervention period (P < 0.01).ConclusionsStretching training and high-intensity interval training appear promising for enhancing BDNF level in neurodegenerative diseases, while isolated aerobic or resistance training show relatively lower potential on improving BDNF. Combined training requires sustained implementation for significant benefits. These findings highlight the importance of tailored exercise prescription for improving BDNF levels.