Objective: To investigate how physical exercise (PE) would affect brain-derived neurotrophic factor (BDNF) in randomized controlled trials (RCTs) of healthy subjects.
Methods:Seven databases (PubMed, Web of Science, Cochrane, Embase, PsycINFO, CINAHL, SPORTDiscus) were searched for RCTs assessing the effects of PE on serum and/or plasma BDNF until December 18, 2021. Meta-analysis was performed by random-effects method with standardized mean difference (SMD) and 95% confidence intervals (CIs). Subgroup analysis and meta-regression analysis were conducted to investigate the potential source of heterogeneity. Trim and fill method, and leave-oneout cross-validation were conducted.Results: Eventually, 21 articles, involving 809 participants, were included in the metaanalysis. Overall, both acute (5 trials, SMD: 1.20, 95% CI: 0.36 to 2.04, p = .005) and long-term (17 trials, SMD: 0.68, 95% CI: 0.27 to 1.08, p = .001) PE had significant positive effects on BDNF levels. Via subgroup analysis, studies of long-term PE with larger sample sizes, female participants, participants older than 60 years, and aerobic exercise contributed to a more pronounced improvement on BDNF levels than that found when all studies were combined.
Conclusion:Both acute and long-term PE had significant positive effects on circulating BDNF in healthy subjects. This review suggests that acute exercise and long-term aerobic exercise are powerful forms of PE to enhance neurotrophic effect, especially for female subjects or subjects over 60 years.