image: A brain age prediction model is constructed by leveraging LightGBM algorithm training on 1425 image-derived phenotypes (IDPs) from T1-weighted brain MRI and chronological age. Features initially undergo tree-based feature importance ranking, where top 50 important features are picked out. Next, supervised distance between each feature is calculated then underwent hierarchy clustering to identify redundant feature groups. After removing redundancy, we visually interpret the final selected subset of features using SHAP technique. To deal with bias, predicted brain age was corrected by linear method. B We first investigate correlations between objectively measured PA and BAG using both nonlinear and linear models. Next, to gain insight into PA and brain structures, we investigate correlations between PA and 1425 IDPs using both nonlinear and linear models. C To verify whether PA and brain health was mediated by BAG, we conducted mediation analysis. Cognitive function and brain disorders were selected as brain health outcomes of interest.
Credit: Chen Han., et al, School of Public Health, Hangzhou Normal University
A new study leveraging accelerometer and brain MRI data reveals that moderate levels of physical activity may help slow brain aging. Led by Associate Professor Chenjie Xu from the School of Public Health at Hangzhou Normal University, in collaboration with institutions including Tianjin University of Traditional Chinese Medicine and Tianjin Medical University, the study is published in Health Data Science.
Analyzing 16,972 participants from the UK Biobank, researchers applied a LightGBM machine learning algorithm to over 1,400 image-derived phenotypes to predict each individual’s "brain age." The findings indicate a U-shaped association between physical activity (PA) intensity and brain age gap (BAG), where both insufficient and excessive PA levels were linked to accelerated brain aging.
Addressing the shortcomings of prior research reliant on self-reported data, this study objectively measured 7-day PA using wrist-worn accelerometers to quantify light (LPA), moderate (MPA), vigorous (VPA), and moderate-to-vigorous (MVPA) activity. Results showed that moderate levels of MPA and VPA significantly reduced BAG (e.g., VPA: β = −0.27), suggesting a brain-protective effect.
Importantly, BAG was found to partially mediate the effects of PA on cognitive function (e.g., reaction time) and brain-related disorders (e.g., dementia, depression). Neuroanatomical analysis revealed that activity-related reductions in BAG were associated with lower white matter hyperintensities and preserved volume in the cingulate cortex, caudate nuclei, and putamen—regions critical for cerebrovascular integrity and cortico-striatal circuitry.
“Our study not only confirms a nonlinear relationship between objectively measured PA and brain aging in a large population, but also provides actionable insight: more exercise isn't always better—moderation is key,” said Xu.
The team’s next step is to build a multi-scale aging framework incorporating sleep, sedentary behavior, neuroimaging, and omics data. Longitudinal studies will investigate how behavioral interventions reshape brain aging, while genome-wide and proteomic analyses aim to uncover the biological mechanisms underlying these effects.
Journal
Health Data Science
Article Title
Accelerometer-Measured Physical Activity and Neuroimaging-Driven Brain Age
Article Publication Date
2-May-2025