Ground Reaction Force Patterns Influence Electromyographic Activity during Stepping in Non-Disabled and Spinal Cord Injured Subjects

Rita U. Lukacs1, Keith E. Gordon2, Janell A. Beres-Jones3, Dan Ferris2,
         Kathy J. Sullivan4 and Susan J. Harkema1,3.
1Brain Research Institute, 3Department of Neurology,
         University of California at Los Angeles, Los Angeles, CA 90095
2Department of Movement Science, University of Michigan, Ann Arbor, MI
4Department of Clinical Biokinesiology and Physical Therapy,
         University of Southern California, Los Angeles, CA


... Introduction

Locomotor training using body weight support on a treadmill (BWST) and manual assistance has emerged as a potential rehabilitation intervention for the recovery of walking following spinal cord injury (SCI) (1-3,10). The success of this approach is dependent on providing appropriate sensory cues to the spinal cord by facilitating the normal kinetics and kinematics of stepping (6,7,9). We previously demonstrated that the type of body weight support (BWS) used to provide the support can affect the result ground reaction forces (GRF) during stepping in non-disabled (ND) and SCI subjects (5). In this study we assessed whether these differences in the pattern of limb loading within a step modulates motor output during stepping in non-disabled and SCI subjects.


... Methods

We studied three ND and three SCI subjects classified by the American Spinal Injury Association (ASIA) Impairment Scale (4,8) (Table 1). Briefly, injury level, years post injury, height, and ASIA impairment scale lower limb motor scores for each subject are summarized in Table 1.


*ASIA A-no sensory or voluntary motor function below the injury level; **ASIA B -some sensory preservation, but no voluntary motor function below the injury level.

We used three different BWS systems that each result in different GRF patterns (5). The Position Control System (Fig.1A) provided BWS by a static cable, which restricted vertical center of mass movement and attenuated the weight acceptance and push off phases of the GRF (Cal Equipment, Long Beach, CA). The Pneumatic Force Control Systems (Fig.1B and 1C) did not restrict vertical center of mass movement and the Open – Loop (Fig.1B) provided BWS using a pneumatic cylinder regulated by input of a constant value (Vigor Equipment, Stevensville, MI). The Closed - Loop Pneumatic Force Control System (Fig. 1C) provided BWS using a pneumatic cylinder with pressure control using a feedback loop using a continuous signal from a force transducer in series with the support cable (Vigor Equipment, Stevensville, MI; Tescom, Elk River, MN).


Figure 1. BWS Systems The subject wears a modified parachute harness (H) attached by carabiners to a steel cable (C). 1A- The subject’s body weight is supported by winding the cable around a winch (w). 1B, 1C - The subject’s body weight is supported by a pneumatic cylinder (P). 1B - Pressure to the cylinder is supplied by a 200 PSI air compressor (A) and manually adjusted by a pneumatic regulator (R). 1C - Pressure to the cylinder is supplied by a 2000 PSI nitrogen tank (N). A Tescom ER3000 dynamic regulator (E) controls the pressure supplied to the cylinder. A load cell (L) records the support force which is sent to the ER3000. The support force is analyzed using Tescom software running on a desk top computer (D). The software compares a set support force to the actual support force. Every 25 ms the ER3000 adjusts the pressure to the cylinder to equalize the set and actual support force.

ND and SCI Subjects stepped on a treadmill (1.2 - 4.0 mph; 0 - 75% BWS) on the Position Control and Open – Loop Pneumatic Force Control BWS systems. A ND subject also stepped on the Closed Loop Pneumatic Force Control System. During stepping using BWST trainers provided SCI subjects with manual assistance at the lower limbs as needed (Figure 2). Hand placement distal to the patella facilitated extension during the stance phase, and hand placement proximal to the ankle facilitated foot clearance during the swing phase. Bungee cords and/or another trainer positioned behind the subject aided in hip stabilization and weight shifting during stepping as needed.

We measured EMG activity bilaterally from the soleus (SOL), medial gastrocnemius (MG), tibialis anterior (TA), medial hamstrings (MH), vastus lateralis (VL), and rectus femoris (RF) using bipolar surface electrodes. We collected EMG at 1000 Hz (Konigsberg Instruments, Pasadena, CA) using a 24-channel hard-wired analog to digital board and a custom written Labview software acquisition program (National Instruments, Austin, TX). Limb load was measured by recording pressure distribution of the foot’s plantar surface using pressure sensor shoe inserts converted to vertical ground reaction force (Novel, St. Paul, MN). An electromagnetic Polhemus sensor placed on the sacrum of the subject was used to record kinetic information about the pelvis movement during locomotion (Skill Technologies, Phoenix, AZ). Vertical movement of the pelvis during gait has been shown to be representative of vertical COM movement during gait. Level of BWS was collected from a force transducer placed in series with the BWS system cable.

For each condition GRF from the left foot were used to determine step cycle times (defined as heel strike to heel strike of the same foot) and stance (heel strike to toe off) and swing durations (toe off to heel strike). Mean GRF during stance, mean load cell value, load cell range and vertical COM displacement during the step cycle were calculated from 8 steps for each condition. Also for each condition rectified and filtered EMG data, GRF and vertical COM movement from 8 consecutive steps were low pass filtered at 6 Hz, normalized to % step cycle, and averaged to get an overall idea of the interactions of all the variables. Averaged waveforms were assumed to represent the typical activity pattern. Coefficients of variation were calculated to scale standard deviations to a mean value throughout the step cycle.


... Results

BWS systems that produce GRF similar to full weight bearing walking, result in more appropriate EMG activity during stepping

GRF patterns are modulated by load, speed, COM movement and type of BWS system used. SOL EMG pattern is modulated by GRF.


TA EMG pattern is modulated by GRF pattern in ND subjects.

Knee muscle EMG amplitude is modulated by GRF pattern in ND subjects.

In bifunctional muscles, GRF pattern modulates EMG activity during the stance phase, but not during the swing phase in ND subjects.

In some cases, SOL EMG pattern mimics the GRF patterns in both ND and SCI subjects

The loss of peak 2 in the GRF causes a significant reduction in EMG amplitude of flexors in SCI subjects.

GRF patterns modulate knee muscle activity in SCI subjects as well.




... Conclusion