Learning From Falling: Perturbation-based Training to Prevent Falls in Older Adults

The investigators will evaluate the difference between a commonly accepted paradigm of balance training (BT) and a more dynamic and task specific form of balance training, perturbation-based training (PBT) in older adults. BT is a key evidenced based strategy for preventing falls in older adults, however it needs to be regular (2hours/week) and long term (\>6moths) while the average effect is only moderate (24%). The reason for the moderate effect on falls is like to be the non-specific stimulus presented during BT. That is, training tends to be quasi-static and slow and largely based on movements described as strength training. However, when an individual loses balance, they are most often required to implement a rapid and dynamic response. Furthermore, while older adults who are unable to recover balance well have generally lower strength, our recent work has demonstrated that it is not their force producing capability that limits them. Rather it is the ability to access moderate levels of muscular force very rapidly and early in the recovery step that differentiates successful versus unsuccessful recovery of balance. Importantly, the ability to produce an effective and rapid recovery step is predicative of avoiding future real world falls. An emerging fall prevention training regimen is perturbation-based training (PBT). PBT involves rapidly disrupting balance requiring the participant to take rapid steps to recover balance. This is commonly achieved on a laboratory treadmill equipped with a safety harness to prevent actual falls during training. By simulating "real-world" balance challenges such as slips and trips, PBT provides a direct means for learning how to recover balance and avoid falls. It has been demonstrated that with only a few PBT sessions, older adults make rapid and dramatic improvements in balance recovery performance, retain the skills long-term and potentially suffer fewer falls over extended periods. This study builds on the previous published work of the PI that describes the key factors related to differences in balance recovery performance, the neuro-motor coordination strategies used during successful and unsuccessful recovery, and currently unpublished pilot studies indicating the efficacy of PBT. To date studies have not directly compared BT regimes recommended by the American College of Sports Medicine (ACSM) against PBT, nor have they evaluated the influence of training on the incidence of real-world falls. In part this may be because PBT currently requires the use of expensive, laboratory treadmills and as such is not accessible by the average independent, community dwelling older adults. A specific randomized study is required and our overall purpose for this study is to compare the balance recovery performance of older adults following either BT or PBT, evaluate differences in the incidence of real-world falls, and develop a safe, effective and portable device for use in future community PBT training studies. The short-term goals are to determine the effect of PBT versus BT and the neuro-motor mechanism of improved recovery behavior. Aim 1: To evaluate differences in balance recovery behavior in older adults following either balance training (BT) and perturbation-based training (PBT) and the incidence on real-world falls. H1: Balance recovery performance will improve in both BT and PBT groups but will be significantly better in those completing PBT when compared to BT. H2: Improvements in balance recovery behavior will be related to improved coordination and neuro-motor control strategies. H3: Real world loss of balance events will be similar in both BT and PBT but incidence of resulting falls will be lower in the PBT group.

APPROACH 1\) The main study The main study to investigate the difference between BT and PBT is a parallel-group randomized controlled trial with 6 months follow-up from baseline to determine the effect of PBT program on stepping behavior and fall incidence in community-dwelling older people. The study has been designed in accordance with the Consolidated Standards of Reporting Trials (CONSORT) statement48. Participants, recruitment and screening Participants will consist of community-dwelling adults aged over 65 years without a recent history of falls (one or more falls in the past 12 months). All participants will be required to obtain medical clearance from their general practitioner prior to participation in the study. Exclusion criteria will be neurological, cardiovascular or musculoskeletal impairment that adversely affect balance or limit involvement in the intervention, diagnosis of dementia, cognitive impairment (score\<24 on Mini Mental State Examination after adjustment for age and years of education49), functional disability/limitation in activities of daily living (Score\>3 on Bayer Activities of Daily Living Scale50), osteoporosis, use of medication that affects balance or causes dizziness (e.g. psychotropics) and participation in supervised exercise one or more times per week. Random allocation and concealment Randomization will be performed following baseline assessment by personnel not involved in recruitment, training or assessments. Participants will be informed that they will receive one of two treatments. Due to the nature of the intervention, research personnel administering the treatments cannot be blinded to group allocation. However, the personnel that perform the prospective falls evaluation will be blinded to group allocation. Intervention BT group. This group will undertake 12 weeks of twice weekly BT in accordance with ACSM guidelines. Training will be conducted in small groups under the supervision of a trainer. Training will involve mobility, resistance and balance training components. PBT group. The PBT protocol uses multi-directional surface translations to provoke rapid compensatory stepping responses to recover balance. The protocol is a modified version of the stepping (but not grasping) component of the PBT program described by Mansfield et al., which was designed to improve rapid stepping reactions and minimize foot collisions in accordance with well-established principles of motor learning including individualization, specificity, progressive overload, and variability of training. At each session the PBT group will experience multi- directional disturbances to standing posture (forwards, backwards, left, right) delivered via translations of a motorized treadmill. 24 perturbations will be administered in a random sequence (with variation in timing, magnitude and direction) and interspersed with 30 s treadmill walking. Successful recoveries during training will be followed by an increase in the degree of difficulty, whereas consecutive failures will result in a return to the perturbation intensity that had previously been performed successfully. Our pilot testing indicates that the prescribed peak accelerations on our motorized treadmill to within ±3% can be achieved. The occurrence of actual falls will be prevented via the use of a custom safety harness attached to an overhead cable instrumented with a force transducer used previously. One familiarization trial in each perturbation direction will be performed at 50% intensity at the beginning of the first training session. Training sessions will be terminated when the participant completes the planned number of trials or if the participant elects not to continue (e.g. feels tired, uncomfortable or unwell). Primary outcome measures The tether-release54 test will be used to determine whether training effects generalize to balance recovery reactions that are not part of the intervention. Loss of balance will be induced by releasing participants from static forward lean postures corresponding to 15%, 20% and 25% of their body weight on a horizontal restraining tether in accordance with the PI's previous work groups published procedures. Participants will be instructed to attempt to regain balance by taking a single, rapid step. Full body kinematics and ground reaction forces will be recorded during all balance recovery tasks using a 3D motion capture system (12-camera Motion Analysis) and force platforms embedded in the ground (AMTI, Watertown, USA). Outcome measures will be margin of stability at touchdown of the stepping leg, step leg kinematics (step length and velocity) and trunk kinematics (trunk angle and trunk angular velocity). All analysis will be performed using OpenSim57 in accordance with the PI's and his previous research units published methods. Secondary outcomes measures Falls frequency will be monitored over the 3 months of the intervention and following 3 months of follow up with monthly falls diaries. Participants will return monthly falls diaries (daily entries) to determine frequency, time, location and cause of falls and related injuries as per consensus recommendations for falls trials52. OVERALL SUMMARY The proposed investigation forms the basis of improved understanding of balance recovery training methods and are important steps for ongoing research into the efficacy of balance recovery training modes relative to reducing the incidence of falls. In particular, the proposed study is the necessary precursor to a large prospective study to evaluate the effect of balance recovery training on health and wellbeing of rural Montanans.