Voluntarily changing step length or step width affects dynamic stability of human walking

Abstract

Changes in step width (SW), step length (SL), and/or the variability of these parameters have been prospectively related to risk of falling. However, it is unknown how voluntary changes in SW and SL directly alter variability and/or dynamic stability of walking. Here, we quantified how variability and dynamic stability of human walking changed when individuals voluntarily manipulated SW and SL. 14 unimpaired, young adults walked on a treadmill at their preferred walking speed with normal gait, with a metronome and with narrower, wider, shorter and longer steps than normal. Taking narrower steps caused increased SL variability while mediolateral (ML) movements of the C7 vertebra (i.e., trunk) became locally more stable (p < 0.05) and anterior–posterior (AP) C7 movements became locally less stable (p < 0.05). Taking wider steps caused increased SW and SL variability, while ML C7 movements became both locally and orbitally less stable (p < 0.05). Any change in SL caused increased SW, SL, and stride time variability. When taking shorter steps, ML C7 movements exhibited greater short-term local and orbital instability, while AP C7 movements exhibited decreased short-term and long-term local instability (p < 0.05). When taking longer steps, AP, ML, and vertical C7 movements all exhibited increased long-term local instability and increased orbital instability (p < 0.05). Correlations between mean SW, SL and dynamic stability of C7 marker motions were weak. However, short-term voluntary changes in SW and SL did significantly alter local and orbital stability of trunk motions.

Introduction

In clinical studies, gait parameters are frequently used to indicate patients’ dynamic stability. However, the correct interpretation of these results is not clear. Walking with wider and shorter steps than normal is often termed “cautious”. However, retrospective studies indicate that individuals who exhibit increased fall risk sometimes walk with shorter [1], [2], longer [3], narrower [1], wider and/or faster [3] steps than normal. Slower walking speeds alone lead to decreased local instability [4], [5], [6] yet are also associated with history of falling [7]. Slower walking speeds also increase motion variability [5], [6], which may or may not also indicate increased risk of falling. In one study, too much or too little step width variability was associated with fall history in older adults who walked at normal speeds (>1 m/s) [8]. Results of these and other retrospective studies do not reveal a clear understanding of the relationship between gait characteristics and fall risk.

Prospective studies on gait characteristics and falling are similarly mixed. Maki [9] found that while shorter and slower steps predicted increased fear of falling, increased variability of stride length and speed doubled an individual's actual likelihood of falling. Additionally, individuals who fell while walking took wider steps with less step width (SW) variability [9]. Contrary to Maki [9], Hausdorff and colleagues [10] found that increased stride time (ST) variability over a 6 min walk predicted falls in older adults. Verghese and colleagues [11] found that slow gait speeds and increases in swing time and stride length variability all predicted increased fall risk in older adults. However, in both studies [10], [11], the greater variability observed could have been due simply to slower walking speeds [5], [6]. DeMott et al. [12] determined that falls in older individuals with peripheral neuropathy were predicted by greater step time variability when walking on irregular surfaces but not on smooth surfaces. Recently, others found that no primary gait variables (means nor variability) predicted falls, but subtle left–right asymmetries in statistical persistence of stride time did [13]. Thus, all five of these prospective studies reached different findings with similar methods and measures.

The fact that shorter, wider and/or more variable steps in patients are associated with increased fall risk might suggest that people who exhibit these patterns are more unstable. However, when external lateral stabilization was applied, young and older subjects both took narrower steps, without changing their mean step length (SL), SW variability, or SL variability [14]. Likewise, when we destabilized healthy young subjects by applying continuous perturbations, they took shorter, wider and faster steps and exhibited greater variability both of stepping parameters and trunk kinematics [15]. These gait changes were accompanied by specific increases in measures of local dynamic instability [16], a measure of within-step dynamic stability. If individuals adopted these gait characteristics to increase their stability, then voluntarily taking shorter or wider steps during unperturbed walking should lead specifically to decreased variability and local instability of stepping parameters and/or trunk movements.

Collectively, these earlier findings suggest two opposing ideas: people increase their risk of falling because they take shorter and/or wider steps, or they take wider and/or shorter steps because they are at greater risk of falling. We wanted to test the latter idea: i.e., that adopting wider and/or shorter steps would decrease individuals’ instability thereby decreasing their risk of falling. We hypothesized that individuals could alter their orbital (i.e., step-to-step) and local dynamic (i.e., within-step) trunk stability by voluntarily changing their SW and SL. We further hypothesized that individuals would exhibit decreased orbital and local instability when walking with wider steps or shorter steps than when walking normally. Our findings would indicate whether and how voluntarily changing gait characteristics contributes to local and orbital stability during walking, the latter of which, in particular, has been linked to fall-risk status [17].