Systematic review and meta-analysis of randomized controlled trials on the effects of exercise interventions on amyloid beta levels in humans.

Alzheimer's disease (AD) represents the most common type of dementia. A crucial mechanism attributed to its development is amyloid beta (Aβ) dynamics dysregulation. The extent to which exercise can modulate this phenomenon is uncertain. The aim of this study was to summarize the existing literature evaluating this issue. A comprehensive systematic search was performed in Pubmed, Scopus, Embase, Web of Science, and SciELO databases and completed in August 2023, aiming to identify randomized controlled trials investigating the effect of exercise upon Aβ-related pathology. The keywords "exercise" and "amyloid beta", as well as all their equivalents and similar terms, were used. For the analysis, the negative or positive dementia status of the subjects was initially considered and then the soluble amyloid precursor protein (sAPP) components and Aβ fragments separately. A meta-analysis was performed and involved eight studies (moderate-to-high quality) and 644 assessments, which were 297 for control and 347 for exercise. No overall effect favoring exercise interventions was observed for both negative (SMD95%=0,286 [-0,131; 0,704]; p = 0,179) or positive AD dementia status (SMD95%=0,110 [-0,155; 0,375]; p = 0,416). The absence of an overall effect favoring exercise interventions was also found for Aβ peptides (SMD95%=0,226 [-0,028; 0,480]; p = 0,081) and for sAPP components (SMD95%=-0,038 50 [-0,472; 0,396]; p = 0,863) levels. Our findings suggest that exercise interventions do not improve Aβ-related pathology in both healthy individuals and individuals with dementia (SMD95%=0,157 [-0,059; 0,373]; p = 0,155), indicating that the beneficial effects of exercise for AD reported in previous studies are related to other mechanistic effects rather than direct amyloid effects (PROSPERO registration number: CRD42023426912).