Elsevier

Gait & Posture

Volume 28, Issue 1, July 2008, Pages 135-143
Gait & Posture

The effect of walking speed on muscle function and mechanical energetics

https://doi.org/10.1016/j.gaitpost.2007.11.004Get rights and content

Abstract

Modulating speed over a large range is important in walking, yet understanding how the neuromotor patterns adapt to the changing energetic demands of different speeds is not well understood. The purpose of this study was to identify functional and energetic adaptations in individual muscles in response to walking at faster steady-state speeds using muscle-actuated forward dynamics simulations. The simulation data were invariant with speed as to whether muscles contributed to trunk support, forward propulsion or leg swing. Trunk support (vertical acceleration) was provided primarily by the hip and knee extensors in early stance and the plantar flexors in late stance, while trunk propulsion (horizontal acceleration) was provided primarily by the soleus and rectus femoris in late stance, and these muscle contributions all systematically increased with speed. The results also highlighted the importance of initiating and controlling leg swing as there was a dramatic increase at the higher walking speeds in iliopsoas muscle work to accelerate the leg in pre- and early swing, and an increase in the biarticular hamstring muscle work to decelerate the leg in late swing. In addition, walking near self-selected speeds (1.2 m/s) improves the utilization of elastic energy storage and recovery in the uniarticular ankle plantar flexors and reduces negative fiber work, when compared to faster or slower speeds. These results provide important insight into the neuromotor mechanisms underlying speed regulation in walking and provide the foundation on which to investigate the influence of walking speed on various neuromotor measures of interest in pathological populations.

Introduction

Recent modeling studies of walking at self-selected speeds have identified how individual muscles work in synergy to satisfy the task demands including body support, forward propulsion and swing initiation (e.g. [1], [2], [3], [4], [5]). These analyses revealed that young adults walking at a self-selected speed utilize a distribution of hip and knee extensor muscle force in early stance and ankle plantar flexor and rectus femoris force in late stance to provide support and forward propulsion. However, how the muscle contributions to these important functional tasks change with walking speed is not well understood. Intuitively, walking at faster steady-state speeds would necessitate an increase in activity for muscles that contribute to forward propulsion. However, faster walking speeds are also associated with longer stride lengths (e.g. [6]), which may require increased activity from those muscles contributing to leg swing (e.g. [7]), and increased activity from those muscles contributing to vertical support because the vertical excursion of the body's center of mass increases (e.g. [8]). Conversely, walking at slower speeds may be mechanically less efficient (e.g. deviating more from natural frequency of the pendular movement so that additional muscular effort may be required) and less conducive to the storage and recovery of elastic energy in the musculotendon complex.

Analysis of muscle activity as walking speed increases has shown that the fundamental phasing relative to regions of the gait cycle remains relatively stable (e.g. [9], [10], [11], [12]). However, walking speed influences each muscle's contractile state (i.e. fiber length and velocity), which may alter the muscle's ability to generate force and power. The potential influence of intrinsic muscle properties on muscle coordination was evident in a recent study showing that the ability of the ankle plantar flexors to produce force as walking speed increased was greatly impaired, despite an increase in muscle excitation, due to sub-optimal contractile conditions (i.e. increased muscle fiber lengths [13]). Since the plantar flexors have been shown to be important contributors to support, forward propulsion and swing initiation during normal walking [1], [2], [3], [4], [5], increased output from other muscle groups would appear necessary to compensate for the decreased plantar flexor output.

Understanding how the neuromotor patterns adapt to the changing energetic demands of increased walking speed is further complicated by the potential increase of elastic energy storage and recovery in tendons (e.g. [14], [15], [16]). Gait kinematics and muscle force requirements change with walking speed (e.g. [17]), which may alter the muscle tendon and fiber kinematics and vary tendon elastic energy storage and recovery. Such variations may partially offset the need for increased active force generation to walk faster. However, the muscles most sensitive to increases in walking speed have not been identified.

The goal of this study was to identify the neuromotor modifications responsible for walking at faster steady-state walking speeds using muscle-actuated forward dynamics simulations that emulate the experimentally collected data of young adults walking at a wide range of walking speeds. The dynamic simulations provide a framework to quantify individual muscle and tendon work and the biomechanical energetic mechanisms executed by each muscle to satisfy the task requirement changes with faster walking speeds. We expected the largest adaptations to occur during the stance phase, as swinging the leg during normal self-selected walking speeds is often assumed to be ballistic (e.g. [18]). However, others have suggested the metabolic cost of leg swing is significant (e.g. [19]) and could become more costly with increased walking speed as the acceleration and deceleration of the swing leg increases. Thus, identifying such functional adaptations by individual muscle groups to increasing walking speed will provide important insight into the neuromotor mechanisms underlying speed regulation in walking and provide the foundation on which to investigate the influence of speed on various neuromotor measures of interest in pathological populations.

Section snippets

Musculoskeletal model

A musculoskeletal model and dynamic optimization framework were used to generate muscle-actuated forward dynamics simulations of walking at 0.4, 0.8, 1.2, 1.6 and 2.0 m/s. The bipedal sagittal plane musculoskeletal model (Fig. 1) and optimization framework used to produce the simulations using optimal tracking have been previously described in detail [3], [4]. Briefly, the musculoskeletal model was generated using SIMM (MusculoGraphics, Inc.) [20] and consisted of rigid segments representing the

Results

The walking simulations at each of the five walking speeds (0.4, 0.8, 1.2, 1.6 and 2.0 m/s) emulated well the group-averaged kinematics and ground reaction forces (e.g. Fig. 1) similar to our previously published work (e.g. [3], [4], [13], [28]). The corresponding muscle excitation patterns also closely mimicked the EMG linear envelopes, which generally increased in magnitude as walking speed increased (Fig. 2). The exceptions were BFsh and GMED, and RF and VAS in late swing, which did not

Discussion

Analysis of individual muscle energetics showed SOL had the largest positive fiber work output among all muscles during stance and it systematically increased with speed (Fig. 4A), which is consistent with its role to provide trunk forward propulsion [3], [5]. GAS and RF work output also increased with speed, which is to be expected because of their synergism to provide trunk forward propulsion [4], [5]. SOL and GAS activity was relatively stable during mid-stance with increasing walking

Acknowledgements

This work was supported by The Whitaker Foundation and NIH grant R01 HD46820. The authors also gratefully acknowledge Felix Zajac for his helpful comments on the manuscript.

Conflict of interest: There is no conflict of interest regarding the publication of this manuscript.

References (36)

Cited by (0)

View full text