Exposure to trips and slips with increasing unpredictability while walking can improve balance recovery responses with minimum predictive gait alterations

Abstract

Introduction: The primary study aim was to determine if repeated exposure to trips and slips with increasing unpredictability while walking can improve balance recovery responses when predictive gait alterations (e.g. slowing down) are minimised. The secondary aim was to determine if predictive gait alterations acquired through exposure to perturbations at a fixed condition would transfer to highly unpredictable conditions.

Methods: Ten young adults were instructed to step on stepping tiles adjusted to their usual step length and to a metronome adjusted to their usual cadence on a 10-m walkway. Participants were exposed to a total of 12 slips, 12 trips and 6 non-perturbed trials in three conditions: 1) right leg fixed location, 2) left leg fixed location and 3) random leg and location. Kinematics during non-perturbed trials and pre- and post-perturbation steps were analysed.

Results: Throughout the three conditions, participants walked with similar gait speed, step length and cadence(p>0.05). Participants' extrapolated centre of mass (XCoM) was anteriorly shifted immediately before slips at the fixed location (p<0.01), but this predictive gait alteration did not transfer to random perturbation locations. Improved balance recovery from trips in the random location was indicated by increased margin of stability and step length during recovery steps (p<0.05). Changes in balance recovery from slips in the random location was shown by reduced backward XCoM displacement and reduced slip speed during recovery steps (p<0.05).

Conclusions: Even in the absence of most predictive gait alterations, balance recovery responses to trips and slips were improved through exposure to repeated unpredictable perturbations. A common predictive gait alteration to lean forward immediately before a slip was not useful when the perturbation location was unpredictable. Training balance recovery with unpredictable perturbations may be beneficial to fall avoidance in everyday life.

Fig 1. The trip and slip perturbation system used for this study.

A slip was induced by a movable tile on two hidden low-friction rails with linear bearings that could be unlocked to slide up to 70cm upon foot contact. A trip was induced using a 14cm height tripping board that could be triggered to spring up from the walkway. The tripping board and slipping tile were not visually detectable and could be moved to any location along the walkway. Black and white vinyl stepping tiles were placed on the walkway to reproduce individual’s step length and a metronome was set to the individual’s cadence.

Fig 2. Gait parameters during non-perturbed trials (N1 to N6, n = 10) interspersed within slip and trip trials throughout the protocol.

The dots and error bars are means and standard errors, respectively. * p < 0.05, ** p < 0.01, n.s. p > 0.05.

Fig 3. Changes in pre- and post-slip step kinematics during slip trials (n = 10).

Margin of stability (MoS) was the distance between an extrapolated (i.e. velocity-corrected) centre of mass (XCoM) to the closest base of support limit at foot touch down. XcoM displacement was the distance between the XCoM to the ankle joint of the supporting limb in the sagittal plane. The dots and error bars are means and standard errors, respectively. The arrows indicate the possible directions relating to better stability. S: slip. * p < 0.05, ** p < 0.01, *** p < 0.001, n.s. p > 0.05.

Fig 4. Changes in pre- and post-trip step kinematics during trip trials (n = 10).

Margin of stability (MoS) was the distance between an extrapolated (i.e. velocity-corrected) centre of mass (XCoM) to the closest base of support limit at foot touch down. XcoM displacement was the distance between the XCoM to the ankle joint of the supporting limb in the sagittal plane. The dots and error bars are means and standard errors, respectively. The arrows indicate the possible directions relating to better stability. T: trip. * p < 0.05, ** p < 0.01, *** p < 0.001, n.s. p > 0.05.

Fig 5

Strategies for balance recovery after slips (A) and trips (B) (n = 10). S: slip, T: trip. Changes in proportions of balance recovery strategies used were examined by applying the generalized linear mixed model (multinomial or binomial logistic regression). A significant effect of trials was observed in the trip trials (p<0.01). This indicate that the proportions of the lowering strategy in response to trips decreased (T1: 90% to T12: 10%) and the elevating-contact (T1: 0% to T12: 50%) and elevating-cross (T1: 10% to T12: 40%) strategies increased. There was no significant effect of trials in the proportion of strategies during the slip trials.