Quantification of spatiotemporal parameter behavior during walking speed transitions
Introduction
Constant speed walking has been rigorously investigated (Bertram and Ruina, 2001, Kadaba et al., 1989, Kirtley et al., 1985, Martin and Schmiedeler, 2014, Srinivasan et al., 2008, Zatsiorky et al., 1994). However, humans often start, stop, and change speeds (Orendurff et al., 2008). These speed transitions have not been studied thoroughly, introducing the question: what biomechanical behaviors do healthy humans exhibit when transitioning between speeds? The fundamental understanding gained by answering this could be applied to develop rehabilitation techniques (Kuo and Donelan, 2010) and assistive gait technologies (Liu et al., 2016, Wang et al., 2011) that restore both steady locomotion and the critical transitions between them.
Speed transitions are characterized by divergence from a pre-transition state and convergence to a post-transition state. Here, the term state refers to the spatiotemporal (step time, length, and speed) behavior exhibited during a walking condition. There are infinitely many combinations of step time and length for a given speed and even more possible spatiotemporal behaviors for a given speed transition. Factors include number of steps; which parameters diverge (only step time, only step length, or both); and mode of convergence (initially overshooting or underestimating the post-transition speed).
In starting, stopping, and walking to running, the transition typically requires two steps to achieve steady locomotion (Bishop et al., 2004, Jian et al., 1993, Segers et al., 2013). However, some studies observed longer transitions (Segers et al., 2006). Differences in experimental control and analysis influence these results. Experimental methodology also dictates which transition aspects can be directly controlled. In both overground walking and treadmill speed ramping, the transition magnitude (i.e. the difference between pre- and post-transition speeds) is not prescribed. As such, little is known regarding the effect of transition magnitude on transition behavior.
For constant speed walking, an optimal step time and length exist that minimizes metabolic cost (Ralston, 1958, Zarrugh et al., 1974). It is generally assumed that humans switch from one energetic optimum to another when performing gait transitions; however, factors beyond energetics such as peak forces, gait constraints, or neurological considerations may also influence transition behavior (Bertram, 2005, Hreljac, 1993, Raynor et al., 2002). Further, humans initially execute reflexive spatiotemporal adjustments before slowly converging to energetic optima in response to sharp speed perturbations (Snaterse et al., 2011), implying that not all transition behavior is exactly optimal. Additionally, humans overcompensate for speed fluctuations (Dingwell et al., 2010) indicating that humans do not maintain one fixed gait behavior. These findings suggest that factors beyond metabolic cost influence transition behavior, and that pre- and post-transition spatiotemporal parameters may differ from constant speed parameters.
This work quantified spatiotemporal behavior during speed transitions on a treadmill. First, we hypothesized that the mean spatiotemporal parameters of the post-transition steady state would be different than those of the corresponding constant speed walking baseline (H1). We expected differences to increase with increasing transition magnitude. Second, we hypothesized that large transitions diverge in both step time and step length more frequently than small transitions (H2). Small transitions may achieve the speed change by only altering step time or length while large transitions alter both. Finally, we hypothesized that transition magnitude affected both number of steps in the transition and mode of convergence (H3). We expected an increasing number of steps to converge as transition magnitude increased, and thus more transitions exhibiting indirect convergence to the post-transition state.
Section snippets
Experimental design
23 healthy adults ages 18 to 62 years old (10 male, 13 female, height 148–184 cm, mass 46.5–90.7 kg, BMI less than 30) participated. Participants had no history of neuromuscular or musculoskeletal impairment. Institutional Review Board approval was received, and subjects provided informed consent.
To compare across subjects of varying sizes, five leg length-normalized speeds (Hof, 1996) were prescribed: 0.40, 0.43, 0.47, 0.50, and 0.53 (approximately 1.18 to 1.58 m/s). This surrounds the typical
Results
This section summarizes the differences between baseline and the post-transition state (H1); the effects of transition magnitude on divergence (H2); steps to converge (H3); and modes of convergence (H3).
Discussion
Baseline and post-transition steady-state means were statistically, but likely not meaningfully, different in step time and length. On average, the difference was related to the signed transition magnitude. However, these specific differences did not hold for all subjects, and the step-to-step variability during constant speed walking exceeded the converged differences. The magnitudes were likely on the order of measurement error. Thus, it is unlikely that subjects converged to meaningfully
Conclusions
When performing speed transitions, subjects on average converge to their equivalent constant speed baseline. Transition magnitude has a significant effect on divergence, with per-subject parameter preference in small transitions. Increasing magnitude is also strongly correlated to increasing number of steps to converge and indirect convergence.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors thank Penn State CTSI for Study Finder for recruiting subjects. This work is supported by the NSF GRFP (Grant No. DGE1255832).
References (30)
- et al.
Multiple walking speed-frequency relations are predicted by constrained optimization
J. Theor. Biol.
(2001) - et al.
The effect of velocity on the strategies used during gait termination
Gait Posture
(2004) - et al.
Papers Trajectory of the body COG and COP during initiation and termination of gait
Gait Posture
(1993) - et al.
Influence of walking speed on gait parameters
J. Biomed. Eng.
(1985) - et al.
Predicting human walking gaits with a simple planar model
J. Biomech.
(2014) - et al.
Are transitions in human gait determined by mechanical, kinetic or energetic factors?
Hum. Mov. Sci.
(2002) - et al.
Spatiotemporal characteristics of the walk-to-run and run-to-walk transition when gradually changing speed
Gait Posture
(2006) Constrained optimization in human walking: Cost minimization and gait plasticity
J. Exp. Biol.
(2005)- et al.
The landscape of movement control in locomotion: Cost, strategy, and solution
Front. Psychol.
(2019) - et al.
Local dynamic stability versus kinematic variability of continuous overground and treadmill walking
J. Biomech. Eng.
(2001)
Do humans optimally exploit redundancy to control step variability in walking?
PLoS Comput. Biol.
Scaling gait data to body size
Gait Posture
Preferred and energetically optimal gait transition speeds in human locomotion
Med. Sci. Sports Exerc.
Measurement of lower extremity kinematics during level walking
J. Orthop. Res.
Cited by (1)
A Kinematic Model to Predict a Continuous Range of Human-Like Walking Speed Transitions
2024, IEEE Transactions on Neural Systems and Rehabilitation Engineering