Capturing Real World Losses Of Balance and Recovery Responses in Older Adults At Risk For Falls


In our early work examining loss of balance (LOB) events in at-home activities, we found that LOBs occur nearly daily. These events occur not only in walking activities (trips, slips) but also in more stationary activities like reaching, bending, and changing position. We have now demonstrated the feasibility of gathering continuous IMU data during real-world activities with wrist voice recorders to designate LOBs and recovery attempts (2). We propose to advance this engineering work by identifying, in community-dwelling older adults at fall risk (n=10), key IMU-derived foot and trunk trajectories underlying LOBs and recovery during two-week monitoring of:

Primary Aim 1: Real-world LOBs, indicated by a wrist voice recorder, to include the context under which these LOBs occurred; and Primary Aim 2: Laboratory-based LOBs, occurring during a set of standard balance and mobility tasks that can trigger LOBs.

Primary Hypothesis: A finite set of foot and trunk movements characterize LOB recovery strategies.
Exploratory Hypothesis: Participants with laboratory-based performance reductions and LOBs are more likely to have more real-world LOBs, particularly during non-walking tasks.

Falls and their consequences place a significant burden on our older population and our health care systems. The primary focus of fall prevention has been on reducing individual (intrinsic) and environmental (extrinsic) risk factors but there is little attention on improving loss of balance recovery strategies that might reduce falls and fall injury. Engineering innovations to enhance fall prevention include the development of fall detection devices and improving quantification of laboratory risk assessments. Our work has focused on the application of expertise in inertial measurement units (IMU sensors) to understand losses of balance (LOBs) and recovery methods. We have now demonstrated the feasibility of gathering continuous IMU data during real-world activities with wrist voice recorders to designate LOBs and recovery attempts. We propose to advance this engineering work by identifying, in community-dwelling older adults at fall risk (n=10), key IMU-derived foot and trunk trajectories underlying LOBs and recovery during two-week monitoring of: Primary Aim 1: Real-world LOBs, indicated by a wrist voice recorder, to include the context under which these LOBs occurred; and Primary Aim 2: Laboratory-based LOBs, occurring during a set of standard balance and mobility tasks that can trigger LOBs. Primary Hypothesis: A finite set of foot and trunk movements characterize LOB recovery strategies. Exploratory Hypothesis: Participants with laboratory-based performance reductions and LOBs are more likely to have more real-world LOBs, particularly during non-walking tasks. The ultimate goal of this pilot is to refine methods and determine effect sizes for a prospective NIH RO1 proposal to follow an at-risk cohort for LOB recovery and fall injury in a real world setting, with a link to lab-based predictive assessments. This Engineering Solutions for Injury Prevention pilot proposal will solidify a UM College of Engineering-UM Medical School link and has implications for future fall injury prevention, a key focus of the UM Injury Center. These LOB profile data can: 1) expand present fall and injury risk assessment to include the quality and safety of performing real world functional activities; and 2) direct customized fall prevention programs to train task-specific balance responses and increased awareness of balance control limitations as well as provide guidance for environmental modifications.