Determining Falls Risk in People with Parkinson's Disease Using Wearable Sensors: A Systematic Review

Area of Science:

• Biomedical Engineering
• Neurology
• Rehabilitation Science

Background:

• Falls are a major concern in Parkinson's disease (PD), with a history of falls being the strongest predictor of future events.
• Current methods for identifying falls risk in PD lack sufficient objective biomarkers.
• There is a need for reliable methods to detect falls risk before falls occur in PD patients.

Purpose of the Study:

• To systematically review the utility of wearable sensors in detecting falls risk in individuals with Parkinson's disease.
• To identify specific sensor-derived features that can differentiate between individuals with PD who fall and those who do not.

Main Methods:

• A systematic literature search was conducted across MEDLINE, EMBASE, Cochrane, and Cinahl databases.
• Included studies were assessed for quality based on sample size, data collection, and statistical analysis clarity.
• Data from 24 eligible articles were extracted and synthesized narratively.

Main Results:

• Twelve studies prospectively measured falls, while others used retrospective data; definitions of "faller" varied.
• Most assessments occurred in clinical settings with diverse sensor placements and mobility tasks.
• Gait variability, stride variability, trunk motion, walking speed, and stride length were common sensor-derived measures distinguishing fallers from non-fallers in PD.

Conclusions:

• Wearable sensor technology shows promise for identifying falls risk in Parkinson's disease.
• Specific gait and motion parameters derived from wearable sensors can serve as potential biomarkers for falls risk.
• Further standardization in sensor use and assessment protocols is needed for clinical application.