Improved screening of fall risk using free-living based accelerometer data.

Area of Science:

• Gerontology
• Biomedical Engineering
• Public Health

Background:

• Falls represent a significant public health concern, particularly among older adults, leading to substantial healthcare costs.
• Current fall risk screening methods have limitations, necessitating more effective and accessible assessment tools.
• Early identification of fall risk enables timely interventions, potentially improving patient outcomes and reducing healthcare expenditures.

Purpose of the Study:

• To introduce and evaluate a novel fall risk screening method using hip-worn accelerometers during free-living conditions.
• To develop and validate techniques for extracting fall risk features from ambulatory activity data.
• To compare the performance of accelerometer-based measures against traditional functional tests and lab-based gait analysis.

Main Methods:

• Utilized data from 1705 older adults in Umea, Sweden, including one week of hip-worn accelerometer recordings.
• Collected gait measurements and performance metrics from three functional tests.
• Incorporated retrospective and prospective fall data, documenting falls within 12 months before and after the study.
• Employed machine learning algorithms to analyze accelerometer data and predict fall risk.

Main Results:

• Accelerometer-based measures demonstrated superior performance in predicting falls compared to existing methods.
• For prospective falls, the accelerometer method achieved a sensitivity of 0.61 and specificity of 0.66.
• For retrospective falls, the method achieved a sensitivity of 0.61 and specificity of 0.68.

Conclusions:

• Hip-worn accelerometers provide a robust and effective tool for assessing fall risk in older adults.
• This technology offers a practical and scalable solution for population-level fall screening.
• Accelerometer-based fall risk prediction can lead to earlier interventions and improved public health outcomes.