Research area
Healthy ageing & mobility
Digital biomarkers of movement-related health in older adults — detecting frailty, mobility decline and fall risk from gait and free-living physical activity, and testing what actually changes them.
Frailty detection
Machine Learning Approach for Frailty Detection in Long-Term Care Using Accelerometer-Measured Gait and Daily Physical Activity: Model Development and Validation Study
X Zheng, Z Zeng, KS van Schooten, Y Yang · JMIR Aging, 8, 2025: e77140
Dynamic gait outcomes — stride length variability, sample entropy and gait symmetry — discriminate frailty more sensitively than spatial-temporal measures such as gait speed.
Factors contributing to differences in physical activity levels in (pre)frail older adults living in rural areas of China
X Zhang, X Zheng, H Hobbelen, B van Munster, T Qian, T Yu, L Feng, C Lamoth · PLOS ONE, 20(11), 2025: e0335607co-author
Factors behind differing physical activity levels among (pre-)frail older adults in rural China.
Sustained and regained robustness are not equivalent: four-year risk of frailty or death in older people
X Zheng, Z Zhou, M Pang, Y Yangready to submit
Accelerometer-Derived Sedentary-Bout Accumulation Patterns and Frailty Severity Among Long-Term Care Residents
X Zheng, Z Zhou, K van Schooten, Y Yangready to submit
Differences in Accelerometer-Measured Daily Physical Activity Patterns Between Pre-Frail and Non-Frail Older Adults
X Zhang, X Zheng, Y Yang, H Hobbelen, B van Munster, L Feng, C Lamothco-authorSubmitted
Intervention & mobility
Effects of pickleball on pre-frailty, physical fitness, 24-hour movement behaviors, and quality of life in older adults: a randomized controlled trial
Z Zeng, X Zheng, CHP Sit, SHS Wong, RKW Sum, Y Yang · International Journal of Nursing Studies, 180, 2026: 105549co-author
After eight weeks of supervised pickleball, 42% of pre-frail participants reached non-frail status, against 8% of controls.
Effects of a Customized Multicomponent Physical Activity Program on Gait Pattern Transitions in Long-Term Care Older Adults: A Cluster Randomized Controlled Trial Using Wearable Sensors and Explainable Machine Learning
X Zheng, Z Zeng, C Ho, Z Zhou, Y Yang · JMIR mHealth and uHealth, 2026Under review
Fall prevention
Effect of Customized 3D-Printed Hip Protectors on User Compliance in Older Adults in Care Homes: Protocol for a Cluster Randomized Controlled Trial
Y Yang, X Zheng, X Song, M Pang, Y Kim, K Chan, C Fungco-authorSubmitted
Effects of Visual Tasks on Kinematic Characteristics during Obstacle Crossing in Older Adults
J Shen, J Pan, X Zheng, H Sit, Y Yangco-authorSubmitted
Safety Without Compromise: Assessing if Wearing Hip Protectors Alters Mobility and Balance in Older Adults Living in Care Homes
H Zhang, X Zheng, Y Yang · 2026co-authorUnder review
Explaining age-related gait
Explaining deep learning models for age-related gait classification based on time series acceleration
X Zheng, M Reneman, E Ottem, C Lamoth · Computers in Biology and Medicine, 184, 2025: 109338
SHAP reveals which parts of an acceleration time series a deep network relies on when classifying age-related gait.
Age-related gait patterns classification using deep learning based on time-series data from one accelerometer
X Zheng, E Wilhelm, M Reneman, E Ottem, C Lamoth · Biomedical Signal Processing and Control, 104, 2025: 107406
Deep learning on raw acceleration from a single accelerometer classifies age-related gait patterns without hand-crafted features.
