AUTHOR=Ouattas Abderrahman , Rasmussen Corbin M. , Hunt Nathaniel H. 

TITLE=Severity of Unconstrained Simultaneous Bilateral Slips: The Impact of Frontal Plane Feet Velocities Relative to the Center of Mass to Classify Slip-Related Falls and Recoveries

JOURNAL=Frontiers in Public Health

VOLUME=10

YEAR=2022

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2022.898161

DOI=10.3389/fpubh.2022.898161

ISSN=2296-2565

ABSTRACT=<p>Targeted interventions to prevent slip-related falls may be informed by specific kinematic factors measured during the reactive response that accurately discriminate recoveries from falls. But reactive responses to diverse slipping conditions during unconstrained simultaneous bilateral slips, which are closely related to real-world slips, are currently unknown. It is challenging to identify these critical kinematic factors due to the wide variety of upper and lower body postural deviations that occur following the slip, which affect stability in both the sagittal and frontal planes. To explore the utility of kinematic measurements from each vertical plane to discriminate slip-related falls from recoveries, we compared the accuracy of four Linear Discriminant Analysis models informed by predetermined sagittal or frontal plane measurements from the lower body (feet velocities relative to the center of mass) or upper body (angular momentum of trunk and arms) during reactive responses after slip initiation. Unconstrained bilateral slips during over-ground walking were repeatedly administered using a wearable device to 10 younger (24.7 ± 3.2 years) and 10 older (72.4 ± 3.9 years) adults while whole-body kinematics were measured using motion capture. Falls (<italic>n</italic> = 20) and recoveries (<italic>n</italic> = 40) were classified by thresholding the dynamic tension forces measured in an overhead harness support system and verified through video observation. Frontal plane measurements of the peak feet velocities relative to the center of mass provided the best classification (classification accuracy = 73.3%), followed by sagittal plane measurements (classification accuracy = 68.3%). Measurements from the lower body resulted in higher accuracy models than those from the upper body, but the accuracy of all models was generally low compared to the null accuracy of 66.7% (i.e., predicting all trials as recoveries). Future work should investigate novel models that include potential interactions between kinematic factors. The performance of lower limb kinematics in the frontal plane in classifying slip-related falls demonstrates the importance of administering unconstrained slips and measuring kinematics outside the sagittal plane.</p>