Estimating fall risk with inertial sensors using gait stability measures that do not require step detection

Area of Science:

• Gerontology
• Biomechanics
• Clinical Biomechanics

Background:

• Falls in older adults pose significant societal and individual burdens.
• Current fall risk assessment questionnaires have limited predictive accuracy.
• Objective, quantitative methods for assessing fall risk are needed for clinical application.

Purpose of the Study:

• To investigate the association between nonlinear measures of trunk kinematics during gait and fall history in older adults.
• To evaluate the potential of multiscale entropy (MSE) and recurrence quantification analysis (RQA) for identifying fall risk.

Main Methods:

• Calculation of nonlinear measures (harmonic ratio, index of harmonicity, MSE, RQA) from trunk accelerations during gait.
• Analysis of trunk kinematic data in a large sample of older adults (42 fallers, 89 non-fallers) aged 50+.
• Assessment of univariate associations between kinematic measures and fall history.

Main Results:

• Multiscale entropy (MSE) and recurrence quantification analysis (RQA) parameters in the anterior-posterior (AP) direction were significantly associated with fall history.
• MSE (scale factor τ = 2) achieved 72.5% correct classification of fallers and non-fallers.
• RQA (maximum length of diagonals) achieved 71% correct classification.

Conclusions:

• Nonlinear analysis of trunk accelerations, specifically MSE and RQA, shows potential for objective fall risk assessment.
• These methods are independent of step detection, reducing potential errors.
• MSE and RQA could be valuable tools for identifying individuals for targeted fall prevention programs.