Preliminary Effects of Aerobic Exercise, Training Dose, and Cardiorespiratory Fitness on Resting Plasma Neurotrophic Factors in Alzheimer’s Dementia in the FIT‐AD Trial

Dereck Salisbury,Danni Li,Michael Todd,Ted K.S. Ng,Fang Yu
DOI: https://doi.org/10.1002/alz.077032
2023-01-01
Abstract:AbstractBackgroundVascular factors are increasingly recognized for their roles in the pathogenesis and progression of Alzheimer’s dementia (AD). Cardiorespiratory fitness (CRF), a physiologic indicator of cardiovascular health, and plasma neurotrophic factors, endogenous soluble proteins in the periphery, have been linked to cognition, brain structure and function, and AD and are modifiable by aerobic exercise (AEx), but their interplay remains poorly understood. The objective of this pilot study was to estimate the effect sizes of 6‐month AEx on resting plasma neurotrophic factors and the relationships of CRF and exercise dose with change in resting plasma neurotrophic factors in AD.MethodThis study was a pilot blood ancillary study of the FIT‐AD Trial that randomized participants to 6‐month AEx or stretching control on a 2:1 allocation ratio. The blood study enrolled 26 participants (AEx Group [n = 18] and stretching control [n = 8]). Resting plasma brain‐derived neurotrophic factor (BDNF), irisin, fibroblast growth factor‐21 (FGF‐21], and insulin‐like growth factor‐1 (IGF‐1), and CRF (peak oxygen consumption) were assessed at baseline, 3 and 6 months. Exercise dose was calculated by the heart rate physical activity score (HRPAS) method.ResultThe sample was 77.6±6.99 years of age and had 15.4±3.00 years of education on average. Sixty‐five percent of the participants were male and 96.2% carried at least one copy of APOE e4. Most 6‐month between‐group effect sizes were small except for irisin (d = ‐0.44), indicating that the irisin levels decreased in the AEx group relative to stretching control. The association between exercise dose and changes in CRF with changes in neurotrophic biomarker were weak (r2 ≤ 0.025).ConclusionThe effects of exercise on BDNF, irisin, IGF‐1, and FGF‐21 may be heterogeneous in AD. Weak associations of exercise dose and CRF changes with changes in resting neurotrophic biomarkers were seen. There were weak associations of exercise dose and CRF changes with changes in resting neurotrophic biomarkers. Our findings need to be validated in future, adequately powered exercise studies in AD.
What problem does this paper attempt to address?