Related Experiment Video
Updated: Jan 9, 2026
08:56
Clinical Assessment of Spatiotemporal Gait Parameters in Patients and Older Adults
Published on: November 7, 2014
14.2K
Accelerometer-Based Gait Analysis as a Predictive Tool for Mild Cognitive Impairment in Older Adults
Junwei Shen1, Yoshiko Nagata2, Toshiya Shimamoto3,4
1Laboratory for Data Sciences, Research and Education Institute for Semiconductors and Informatics, Kumamoto University, 2-39-1, Kurokami, Chuo-ku, Kumamoto 860-8555, Japan.
Sensors (Basel, Switzerland)
|December 11, 2025
Summary
Accelerometer gait analysis shows promise for detecting cognitive decline in older adults. This non-invasive method uses walking patterns to identify early signs of cognitive impairment, aiding in timely diagnosis.
Area of Science:
- Gerontology
- Neurology
- Biomedical Engineering
Background:
- Cognitive impairment, including dementia, poses a significant health challenge in aging populations.
- Early detection of cognitive decline is crucial for effective management and intervention.
- Current diagnostic methods can be invasive or require specialized clinical settings.
Purpose of the Study:
- To investigate the efficacy of accelerometer-based gait analysis for non-invasive prediction of cognitive impairment in older adults.
- To evaluate the performance of machine learning models using gait-derived features for cognitive status classification.
- To explore the potential of gait dynamics as a biomarker for early cognitive decline.
Main Methods:
- Collected gait data from 75 older adults (cognitively normal and with dementia) using a waist-worn accelerometer during self-paced walking.
- Applied Allan variance (AVAR) to gait data to extract frequency stability features.
- Utilized logistic regression and Light Gradient Boosting Machine (LightGBM) models with AVAR features and age to classify cognitive status (based on Mini-Mental State Examination scores).
Main Results:
- LightGBM model achieved a high area under the curve (AUC) of 0.92 for cognitive status classification, outperforming logistic regression (AUC = 0.85).
- Gait analysis indicated that individuals with mild cognitive impairment (MCI) exhibited gait patterns closer to those with dementia, even when grouped with cognitively normal participants.
- AVAR-derived gait features demonstrated significant potential in distinguishing between different cognitive states.
Conclusions:
- Accelerometer-based gait analysis using AVAR features is a promising non-invasive tool for predicting cognitive impairment in older adults.
- This approach may facilitate earlier detection of cognitive decline, including MCI.
- Gait dynamics captured by accelerometers could serve as a valuable digital biomarker in gerontology and neurology.
