Neural modulation of isometric contractions

MISSING IMAGE

Material Information

Title:
Neural modulation of isometric contractions
Physical Description:
ix, 187 leaves : ill. ; 29 cm.
Language:
English
Creator:
Robichaud, Julie A., 1964-
Publication Date:

Subjects

Subjects / Keywords:
Exercise and Sport Sciences thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Exercise and Sport Sciences -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1997.
Bibliography:
Includes bibliographical references (leaves 177-187).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Julie A. Robichaud.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 37819142
System ID:
AA00024933:00001

Table of Contents
    Title Page
        Page i
        Page ii
    Acknowledgement
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
        Page vii
    Abstract
        Page viii
        Page ix
    Chapter 1. Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
    Chapter 2. Review of literature
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
    Chapter 3. Methods
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
    Chapter 4. Results
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
    Chapter 5. Discussion, summary, conclusions and implications for future research
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
    Appendix A. Informed consent form
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
    Appendix B. IRB form
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
    References
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
        Page 186
    Biographical sketch
        Page 187
        Page 188
        Page 189
Full Text











NEURAL MODULATION OF ISOMETRIC CONTRACTIONS











By

JULIE A. ROBICHAUD














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1997
































Copyright 1997

by

Julie A. Robichaud













ACKNOWLEDGMENTS

I would like to take this opportunity to thank those who made my doctoral education special and this dissertation possible. I would like to express my gratitude to my dissertation committee members Dr. Keith Tennant, Dr. Denis Brunt, Dr. Mark Trimble, and Dr. Robert Singer, for their guidance and support throughout the dissertation process. In particular, I appreciate Dr. Tennant, who believed in me.

A very special thank you goes to Dr. Denis Brunt for his mentorship, training, and nurturing. He has left me with a lasting impression of the way to conduct research in the future. I would like to recognize the Educational Division of the American Physical Therapy Association for its financial support of my dissertation work. Additionally, I would like to express my gratitude to my friends, fellow students, and professional colleagues who provided me with technical assistance and emotional support. In particular, I want to thank Bill McClancy, a special friend who was always there to help me out of any technical jam. I would also like to thank the physical therapy staff at Tacachale, who were always willing to accommodate my fluctuating work schedule.

I would also like to express my love and affection to my parents, Kathy and Robert Robichaud, and my brother, Robert.





iii








They were always there to give me encouragement and love when I needed it most. words only fail in showing my gratitude.

















































iv















TABLE OF CONTENTS




ACKNOWLEDGMENTS ............................................ ii

ABSTRACT ................................................. viii

CHAPTERS

1 INTRODUCTION ...................................... 1

Single-Joint Movements ............................ 4
Dual-Strategy Hypothesis .......................... 9
EMG Activity and Torque ..................... 11
Isometric Contractions ...................... 12
Excitation Pulse ............................ 14
Motor Neuron Reflex Excitability ................. 16
Statement of Problem ............................. 20
Research Hypothesis .............................. 21
Study 1 ..................................... 21
Study 2 ..................................... 23
operational Definitions ....... .................. 26
Basic Assumptions ................................ 28
Limitations ...................................... 29
Significance of the study ........................ 30

2 REVIEW OF LITERATURE .............................. 33

Motor Programs ................................... 36
Models of Programmed Movement .................... 40
Fitt's model ................................ 44
impulse Timing model ........................ 47
Pulse Height model .......................... 50
Dual-Strateqy Model ......................... 52
Speed-insensitive strategy ............. 54
Speed-sensitive strategy ............... 56
Excitation Pulse ................................. 57
Isometric Contraction ............................ 58
Electromyographic Activity ....................... 61
motor Neuron Pool Excitability .............. 65
Motor Neuron Excitability and Force ......... 69
Reaction Time Fragmentation ...................... 70
Spinal Facilitation ......................... 71
Motor Neuron Reflex Excitability (Pulse-Width
vs Pulse-Height) .......................... 73



v









H-Reflex Facilitation, Force
and Reaction Times ........................ 76
Summary ......................................... 77

3 METHODS .......................................... 78

Participants ..................................... 78
Selection of Force Levels and Bandwidths ......... 79 Instrumentation .................................. 82
EMG ......................................... 82
Stimulation ................................. 83
Force Transducer ............................ 83
Participant Positioning .......................... 84
Testing sessions ................................. 84
Procedures ....................................... 86
Study 1 ..................................... 86
Study 2 ..................................... 88
Design and Analysis .............................. 89
Study 1 ..................................... 89
Study 2 ..................................... 90

4 RESULTS .......................................... 93

Study I .......................................... 94
Temporal Measures ........................... 94
Slope of dF/dt ant the Slopes of Force ...... 97
Peak of the First Time Derivative of Force
(dF/dt) ................................... 103
Soleus EMG Activity (Q30 and Qacc) ......... 107
Summary of Results Study 1 ................... 109
Same Force Level Different Bandwidths .... 109 Different Force Levels Same Bandwidth .... 111
Study 2 ...... .................................. 112
Time of Facilitation ....................... 113
Slope of Facilitation ...................... 120
Peak Facilitation .......................... 121
Summary of Results .............................. 125

5 DISCUSSION, SUMMARY, CONCLUSIONS
AND IMPLICATIONS FOR FUTURE RESEARCH ............ 126

Effect of Bandwidth Changes on Movement
Trajectories and EMG Parameters ................ 127
Effect of Force Changes on Movement
Trajectories and EMG Parameters ................ 128
Neural Excitation Pulse ......................... 131
Movement Accuracy .......................... 132
Force Level ................................ 133
Task Parameters ............................ 134
Models of Programmed Movement and Neural
Control ........................................ 138
Pulse-Height Vs. Pulse-Width Control ....... 138 Fragmentation of Premotor Time ............. 142
ordering of H-reflex Trials ................ 144


vi









Summary ......................................... 146
Conclusions ..................................... 146
Implications for Future Research ................ 148

APPENDICES

A INFORMED CONSENT FORM ........................... 152

B IRB FORM ........................................ 159

REFERENCES ................................................ 177

BIOGRAPHICAL SKETCH ....................................... 187













































vii














Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

NEURAL MODULATION OF ISOMETRIC CONTRACTIONS

By

Julie A. Robichaud

August, 1997


Chair: L. Keith Tennant
Cochair: Denis Brunt
Major Department: Exercise and Sport Sciences


The dual-strategy hypothesis was developed to explain how single-joint voluntary movements were controlled. Movements were divided into two different strategies which were suggested to be modulated by different excitation pulses. Changes in the H-reflex prior to the movement may be one way to evaluate the existence of this excitation pulse. This study was designed to test the existence of the excitation pulse by quantifying changes in spinal excitability that may occur with changes in bandwidth and force level during a ballistic ankle plantar flexion isometric contraction. Eleven participants were tested and each was seated in a modified chair with their hips at 900, knees at 120% and the ankles neutral. A force transducer was placed to measure ankle force output. Following a visual stimulus subjects were trained to produce a plantar flexion



Viii















force of 25% and 50% of a maximum voluntary contraction, within bandwidths of 5% and 15% of the selected force level. Soleus motor neuron reflex excitability was analyzed measuring changes in the H/M ratio. The H-reflex was randomly elicited at intervals of 50, 75, 100, 150, and 175 ms following the visual stimulus. A two-way repeatedmeasures analysis of variance indicated a significant effect among bandwidths for the time of change in spinal excitability = 21.2; p




ix















CHAPTER 1
INTRODUCTION


Individuals perform hundreds of voluntary single and

multijoint movements during the course of a single day. when observing these movements, there is enormous diversity. There are a number of factors, including the angle of movement, number of joints involved in the movement, body position of the individual, and environmental factors (i.e., wind, rain, ect.), which can influence how a movement is performed. Nonetheless, these movements seem to produce some comnmonalties, which may be regarded as movement regularities. The neurophysiological and musculoskeletal mechanisms involved with multi- and single-joint voluntary movements may be difficult to understand. The underlying mechanisms (neurological and musculoskeletal), which may control these regularities of voluntary single-joint movement have been investigated through, using a variety of experimental methods (Ghez & Gordon, 1987; Gottlieb, Corcos, & Agarwall, 1989a; Schmidt, 1979). When multi- and single-joint movements are understood in healthy individuals, this may lead to a better understanding of movement dysfunction in patient populations.

Movement dysfunction in patient populations such as

those who have Parkinson disease or have sustained a stroke is not well understood. The most appropriate method to use



1






2




for motor relearning or rehabilitation is difficult to asses unless researchers are able to identify the precise nature of the dysfunctions that produce cognitive, programming, or execution deficits. The analysis of movement dysfunction in patients would be facilitated if there were a comprehensive model for single-joint control. This may enhance the practitioner's ability to identify which patients have cognitive, programming, or execution deficits that impede the desired movement. A model, that encompasses both programming and execution components of movement, would provide a more fundamental basis for the understanding of movement dysfunction. Therefore, therapists, abilities to focus treatments on the actual movement deficits might be enhanced.

A model of single-joint control, the dual-strategy

hypothesis, has been recently developed by Gottlieb, Corcos, and Agarwall (1989a). Supportive research has established that predetermined movement strategies can be differentiated according to specific electromyographic (EMG) activity and torque trajectories and may be divided into speed sensitive

(SS) and speed insensitive (SI) strategies. The SS strategy is adopted when the speed of the movement is implicitly or explicitly controlled. In the SI strategy, there is no intention or requirement to control movement speed. This model of single-joint control has been deemed a satisfactory explanation of isotonic movements; however, ambiguity still






3



exists as to whether isometric contractions may be appropriately classified using this model.

The utilization of the SS or SI strategy is based upon different excitation pulses with the same initial pulse determining both force trajectories and EMG activity. The SS and SI strategies should elicit different excitation pulses. These excitation pulses may be represented by changes in spinal cord excitability prior to movement onset, with the different strategies producing different excitation pulse patterns (Gottlieb, Corcos, & Agarwall, 1989a). The SS strategy should modulate the excitation pulse height, while the SI strategy may alter the width of the excitation pulse. Although Gottlieb et al. (1989a) outlined specific relationships between torque trajectories and agonist EMG activity with different movement strategies, there was no attempt to evaluate the existence of the excitation pulse or to show relationships between this pulse and motor programming.

Evidence suggests different excitation pulse patterns may be reflected by different patterns of movement. For example, in an anticipation timing task, which could be classified as an SI strategy, Frank's (1976) data showed an increase in the time period of motor neuron excitability. This would be classified as pulse width modulation. Eichenberger and Ruegg (1984) ascertained that faster reaction times were associated with higher levels of motor






4



neuron excitability that would be classified as pulse height modulation. In this study, an attempt will be made to analyze the relationships between torque trajectories and agonist EMG activity of single-joint isometric movements. Additionally, the proposed excitation pulse will be evaluated by measuring changes in spinal excitability prior to movement onset during voluntary isometric movements.



Single-Joint Movements



Woodworth (1899) was one of the first researchers to

analyze single-joint voluntary movement. Woodworth indicated that two phases of movement controlled voluntary single-joint limb movements. Since Woodworth's original research, many investigators have analyzed various combinations of limb trajectory, ENG amplitude and duration, movement time, acceleration, and peak velocity data while attempting to determine the existence of strategies for performing singlejoint movements (Corcos, Gottlieb, & Agarwall, 1989; Gottlieb, Corcos, & Agarwall, 1989a, 1989b; Fitts, 1954; Gordon, & Ghez, 1987a; Schmidt, 1976). A strategy consists of a set of rules that may be defined as consistent relationships between two parameters of movement (Gottlieb, Corcos, & Agarwall, 1989a).

Fluctuating relationships among dependent and

independent variables have confounded the issue of the






5



relationships between variables of single-joint movements (Gottlieb, Corcos, & Agarwall, 1989a). Rules may exist among both dependent and independent variables. A dependent variable may include movement time, velocity, EMG amplitude, EMG burst duration, or errors. Rules between dependent variables demonstrate a relationship that is not directional and usually not causal. That is, peak agonist EMG activity and peak velocity (dependent variables) are both associated with distance (independent variable) moved. However, each dependent variable can independently increase or decrease with different distances. Independent variables are under control of the experimenter and may include manipulation of target size, movement distance, movement time or inertial load (Gottlieb, Corcos, & Agarwall, 1989a). Rules existing between dependent and independent variables have also been identified. This is observed in relationships between movements to various distances and with different inertial loads. For example, as target distance increases (independent variable) both movement time and velocity increase (dependent variables), while an increase in target size (independent variable) results in a decrease in movement time with a corresponding increase in velocity (dependent variable) (Fitts, 1954). Additionally, based upon the experimental design, a dependent variable may become an independent variable. For example, if movement time is






6




constrained, then movement time becomes an independent variable.

There is no comprehensive model of general rules for governing single-joint movement. Some models explore descriptive relationships between kinetic and kinematic variables, EMG activity and trajectory control, or movement speed and accuracy, and are therefore limited to specific independent and dependent variables (Corcos, Gottlieb, & Agarwall, 1989, Fitts, 1954; Gordon & Ghez, 1987a; Gottlieb, Corcos, & Agarwall, 1989a, 1989b; Schmidt, 1988). Theoretically, a more comprehensive model of single-joint movement should predict multiple relationships among dependent and independent variables. A comprehensive model should provide a basis for the change in variable relationships in a variety of experimental paradigms. In addition, the model should predict the multiple variable relationships prior to the actual movement.

Some researchers have developed models which explain single-joint control by observing and then describing the movement based on a mathematical modeling of these observations (Newell & Corcos, 1993; Schmidt, 1979). Some of the basic models that describe relationships among dependent and independent variables include the impulse-timing model (Schmidt, 1979; Schmidt, Zelaznik, Hawkin, Frank, & Quinn, 1979), Fitts' (1954) speed-accuracy model, and Gordon and Ghez's (1987b) pulse-height model. The impulse-timing model






7



proposes that a descending impulse controls movement. This model assumes that a motor program controls the amplitude and duration of the isometric force impulse, which in turn regulates force output. Force output is presumably controlled by either the initial burst of force or the time over which the force is transmitted. when either variable, the initial burst of force or force duration is controlled, the other variable is proportionally recalled (Schmidt, 1988; Schmidt et al., 1979).

The speed-accuracy model describes the relationship

between the speed of movement and target size. This model is based upon a logarithmic trade-off between movement speed and accuracy, with the trade-off relating to the accuracy component of the task. Therefore, this model accounts for adjustments in movement speed with corresponding changes in target size (Fitts, 1954; Newell & Corcos, 1993; Schmidt, 1988). This model appears to be limited to specific experimental paradigms and can only explain limited relationships among dependent and independent variables.

Speed-accuracy and impulse timing models have been associated with relationships between dependent and independent variables as a means of determining single-joint voluntary control. The pulse-height model expanded on these models with the addition of an instructional variable, as another independent variable, which could influence singlejoint control (Gordon & Ghez, 1987b). Although instructions






8



were provided in the previous models, the investigators did not classify and analyze instructions as an independent variable. Instructional variables are premovement instructions that are given to the participant by the experimenter; e.g., to execute the task within a certain time frame, as fast as possible, or as accurately as possible.

Developed for isometric contractions is the pulse-height model which suggests that force impulses are generated by rules governing pulse-height control (Gordon & Ghez, 1987b). That is, force impulses are proportionally scaled to allow responses of different amplitudes. The force impulse controls the rate of rise of force, while the control of force rise time depends upon the independent variable. In an accuracy task, force rise time is considered to be invariant. That is, force rise time presumably will not increase with increased peak force. However, in tasks in which there are no accuracy constraints, force rise time will increase with peak force. This model (pulse-height) is based upon the aimed force impulse which subsequently determines the trajectory of motion. The dual-strategy model expands upon the pulse-height model by providing further explanation as to the origin of control for single-joint movements.


Dual-Strategy Hypothesis


Gottlieb, Corcos, and Agarwall (1989b) developed the dual-strategy model based on the analysis of isotonic






9



(concentric) movements as an alternate model for the control of single-joint voluntary movements. This model is dependent upon the concept of an excitation pulse, which activates the spinal motor neuron pool. There is no delineation as to the origin of the excitation pulse, but parameters that may influence the modulation of the excitation pulse are defined. The excitation pulse is used as an explanation for limb trajectories (peak force and rate of rise of force development) and EMG activity that could be affected by task changes (Gottlieb, Corcos, & Agarwall, 1989a, 1989b; Newell & Corcos, 1993). This model is similar to the pulse-height model with two exceptions. The pulse height model did not interrelate limb trajectory with EMG activity; nor does it provide an explanation for the origin of the isometric force impulse. Therefore, it appears the dual-strategy model better defines the programmed aspects of movement, while accounting for relationships among dependent and independent variables. Additionally, this model predicts the multiple variable relationships prior to the movement (Gottlieb, Corcos, & Agarwall, 1989a).

The dual-strategy hypothesis can account for isotonic movements across different distances, inertial loads, different target widths, and over different velocities (Corcos, Gottlieb, & Agarwall, 1989; Gottlieb, Corcos, & Agarwall, 1989a, 1989b). The dual-strategy model is based upon three propositions: 1) movements are planned according






10



to strategies which are a set of rules among independent variables, 2) the control of single-joint movements is based on at least two strategies which include SI and SS strategies, and 3) the choice between the strategies depend on the independent variable as to whether to control the speed of the movement (Gottlieb, Corcos, & Agarwall, 1989a).

Gottlieb, Corcos, and Agarwall (1989a) define the SI

strategy as consisting of movements that are "insensitive', to the speed of the movement. Single-joint movements performed under this strategy lead to a linear relationships among peak accelerating torque and peak torque, peak velocity, and movement time. Additionally, when these variables are plotted against time, all initially rise along the same trajectory. These relationships still hold when independent variables, such as distance moved or inertial load, are changed. For example, as distance decreases, there is a linear decrease in peak velocity, peak torque, and movement time (Gottlieb, Corcos, & Agarwall, 1989a, 1989b).

The S5 strategy consists of movements where speed is the independent variable, which may be implied either explicitly or implicitly. For example, participants are instructed to move within a given time period, to different sized targets at the same distance or to move as rapidly as possible. In the SS strategy, peak accelerating torques, and peak velocity increase linearly with a corresponding increase in movement speed. However, these measures initially rise along






11



different trajectories Additionally, the SS and SI movement strategies each produce a characteristic pattern of muscular activity and torque profile.



EMG Activity and Torque


It has been implied those relationships between peak

accelerating torque and EMG activity account for differences between the SS and SI strategies (Corcos, Gottlieb, & Agarwall, 1989; Gottlieb, Corcos, & Agarwall, 1989b). The excitation pulse, in addition to controlling limb trajectory, should also modulate agonist EMG activity. To further analyze EMG activity, agonist activity was divided into two separate time periods. The first time period (Q30) consisted of the first 30 ms of agonist activity. The agonist activity following the initial 30 ms until the end of force acceleration was described as Q acceleration (Q,) (Corcos, Gottlieb, & Agarwall, 1989). Depending upon the movement strategy, there were different relationships between EMG activity and peak accelerating torque in isotonic movements.

In the SI strategy, these two phases of EMG activity (Q30 and Q..) were represented by different relationships with peak accelerating torque. The variable, Q30, was shown to be insensitive to the distance moved or inertial load. That is, Q,0 did not appreciably change with the manipulation of movement distance or load (Gottlieb, Corcos, & Agarwall,






12



1989b). In contrast, there was a linear relationship between Q.. and changes in the independent variables (Gottlieb, Corcos, & Agarwall, 1989b). Additionally, the SI strategy was ascertained to be the default strategy for movements where speed was not the independent variable (Gottlieb, Corcos, Agarwall, & Latash, 1990).

With the SS strategy, peak accelerating torque, Q30, and Q,, increased linearly with an increase in movement speed. For example, as the speed of the movement increased there was a corresponding increase between peak accelerating torque and Qa,, This relationship was established for Q30 under the SS strategy (Corcos, Gottlieb, & Agarwall, 1989). Therefore, both Q,0 and Qa= were sensitive to the speed of movement. Additionally, the onset of antagonist EMG activity was delayed with increased movement times (Corcos, Gottlieb, & Agarwall, 1989).



Isometric Contractions


Corcos, Agarwall, Flaherty, and Gottlieb (1990) extended the dual-strategy model of single-joint control to describe isometric contractions. They proposed that isometric contractions are controlled in the same manner as singlejoint isotonic movements, but were unable to fully support this assumption. However, Gordon and Ghez (1987) appear to have support for this assumption, because their evaluation of






13



fast and accurate isometric contractions determined that the first measurable features of a response could predict the entire response. In this case, both the first and second time derivatives of force were used to predict the entire response. In the evaluation of isometric contractions by Corcos et al. (1990), movements performed under the SS strategy appear to support the dual-strategy hypothesis; however, ambiguity exists with the analysis of contractions performed under the SI strategy.

An analysis of the research on isometric movements by Corcos et al. (1990), appears to reveal some limitations in the design and subsequent analysis of the data. Gottlieb, Corcos, and Agarwall (1989b) previously stated that peak accelerating torque is the linking variable that could explain correlations between dependent and independent variables. Isotonic movements yield specific relationships between peak accelerating torque and EMG activity (Q30 and Qacc) with the manipulation of independent variables, which can

differentiate the two strategies (Corcos, Gottlieb, & Agarwall, 1989; Gottlieb, Corcos, & Agarwall, 1989b). For example, with the SS strategy, peak accelerating torque and both Q3. and Qac. increase linearly with an increase in movement speed. However, under the isometric condition, conclusions could not be drawn about the relationships of torque measures and ENG activity. Additionally, there appear






14



to be some questions regarding the classification of the SI and SS tasks.

In a study by Corcos et al. (1989), elbow flexion and

extension movements were performed to varying bandwidths with the same amount of force production. This was classified as an SI strategy. However, the participants were also asked to perform these movements at a comfortable speed. This added instruction made this task a combination of both strategies, which may have been a factor that influenced the lack of expected results for the different strategies. It appears that further research is required to determine if isometric contractions operate under a dual-strategy hypothesis.



Excitation Pulse


The dual-strategy model is based on EMG activity and torque trajectories being the consequences of the same initial control signal. The proposed control signal, the excitation pulse, should account for the different patterns of EMG activity and muscle torque trajectories that are dependent upon the movement strategy. Gottlieb, Corcos, and Agarwall (1989a) proposed that the excitation pulse could be controlled by altering either the pulse's height or width. This implied that two different mechanisms control the excitation pulse, which should correspond with the two different strategies of movement.






15




The excitation pulse is thought to have a uniform intensity when movements are carried out under the SI strategy. This intensity should not be influenced by task variables, such as moving different distances or loads. However, the width of the excitation pulse should vary in proportion to changes in the independent variables. This is described as pulse-width modulation (Gottlieb, Corcos, & Agarwall, 1989b). In contrast, the width of the excitation pulse should be held constant, while the amplitude of the excitation pulse should change in the SS strategy. Therefore, the intensity of the excitation pulse should be influenced by moving at different speeds. This is known as pulse-height modulation (Corcos, Gottlieb, & Agarwall, 1989).

According to the dual-strategy hypothesis, different movement strategies should produce different movement profiles as a consequence of changes in the excitation pulse. According to Gottlieb, Corcos, and Agarwall (1989a), the excitation pulse is reflected by the status of the excitability of the spinal motor neuron pool. The level of motor neuron excitability can be analyzed using the Hoffman reflex (H-reflex) (Hugon, 1973; Schieppati, 1987). Frank (1986) suggested that motor neuron pool excitability relates to the force and timing of the movement. Butler, Yue, and Darling (1993) indicated that the amplitude of the H-reflex linearly increases with force. This increase occurs up to 60% of a maximum voluntary contraction. After 60% of a






16



maximum voluntary contraction, the H-reflex increases in curvilinear fashion. Since the excitation pulse may be predicted from inspection of EMG activity and limb trajectory, the relationships between EMG activity and limb trajectory may be compromised when force output exceeds 60% of a maximum voluntary contraction. In fact, some of the tasks in the Corcos et al. (1990) isometric study exceeded 60% of a maximum voluntary contraction. This may be a plausible explanation for the lack of conclusive results. Further research seems necessary to evaluate the potential relationships among the excitation pulse, EMG activity, and peak torque while isometric contractions are performed. Additionally, the existence of an excitation pulse and its potential relationship with spinal excitability warrants investigation. This will help to determine if different movement strategies are mediated at a spinal cord level.



Motor Neuron Reflex Excitability



Motor neuron pool excitability increases prior to the generation of plantar flexion torque (Brunt & Robichaud, 1996; Kagamihara, Komiyama, Ohi, & Tanaka, 1990; Mitchie, Clarke, Sinden, & Glue, 1976; Sullivan, 1980). This increase occurs 50 to 80 ms prior to ENG activity onset in a nonchoice ballistic task and is time-locked to the onset of the EMG activity (Brunt & Robichaud, 1996; Kots, 1977; Mitchie et






17



al., 1976). This period of spinal cord facilitation implies the readiness of the central nervous system to implement the planned motor action (Brooks, 1986; Frank, 1986). If the excitation pulse is reflected in changes in spinal excitability, then changes in the period of spinal facilitation should correspond with changes in movement strategies that determine EMG and torque profiles.

There appears to be support for different movement strategies being reflected by changes in spinal cord excitability prior to onset of EMG activity. The dualstrategy hypothesis suggests that the SI strategy should be reflected in changes in the timing (pulse width) of spinal cord facilitation. Kagamihara et al. (1990) evaluated the time course of spinal facilitation in a ramp plantar flexion task. For the ramp task, the participant moved at a preselected cursor speed. The onset of spinal facilitation began at 92 ms prior to EMG onset. Additionally, Frank (1986) ascertained an increase in the time period of spinal cord facilitation (70 ms prior to EMG onset) with an anticipation timing task. The dual-strategy hypothesis proposes that the excitation pulse in an anticipation timing or ramp task should be classified as representing a SI strategy, since there is no requirement to control speed for either of these tasks. These examples appear to support pulse-width modulation of spinal cord facilitation, since the






18



onset of spinal facilitation occurs later than 50 mns prior to EMG activity onset.

Because speed is the dominate feature of ballistic and step tasks, these movements would be classified as representing the use of an SS strategy according to the dualstrategy hypothesis and should show pulse-height modulation. Numerous studies have shown that the onset of spinal facilitation occurs between 55 and 80 mis prior to EMG onset in ballistic plantar flexion tasks (Brunt & Robichaud, 1996; Eichenberger & Ruegg, 1984; Manning & Hammond, 1990). Kagamihara et al. (1990) evaluated spinal facilitation in step and ramp plantar flexion tasks. In the step task, participants were requested to reach a target that was designated between 20 and 40% of the a maximum voluntary contraction. The step task showed an earlier onset of spinal facilitation (55 mns prior to EI4G onset) when compared to the ramp task (92 mns prior to EMG onset). These results appear to support the pulse-height control for SS strategies.

Additionally, Eichenberger and Ruegg (1984) ascertained that shorter reaction times had greater amounts of spinal facilitation. This could be attributed to these tasks being carried out using an SS strategy, while longer reaction times may have defaulted to the SI strategy. Eichenberger and Ruegg (1984) implied that different movement strategies may have been used to perform the tasks. However, none of these studies were designed to analyze SS or SI strategies (e.g.,






19



pulse width or pulse height modulation) and modulation of spinal facilitation.

Single-joint isotonic movements can be classified

according to the dual-strategy hypothesis. There is some ambiguity as to the ability of this hypothesis to be extended to isometric movements. The dual-strategy hypothesis asserts that single-joint movements can be classified into different strategies depending upon whether or not the speed of the movement is controlled. Different movement strategies may be determined by the type of neural excitation pulse. This excitation pulse presumably modulates spinal excitability that should reflect changes in the H-reflex.

The dual-strategy hypothesis appears to provide a model for unifying principles for the control of isotonic singlejoint movement. Isometric movements have been evaluated to determine if these movements are governed by the same principles of single-joint movement. Research on isometric movements has not fully supported the existence of the dualstrategy model. However, an analysis point to limitations in the design of studies oriented toward determining this model's extension to isometric movements. This study is designed to examine isometric movements to see if the dualstrategy hypothesis can be extended to account for isometric movements. The potential neural mechanism (excitation pulse) that may govern the different strategies will also be evaluated.






20



Statement of Problem



The purpose of this study is to examine the relationship between EMG activity and torque trajectories during isometric contractions to determine if the dual-strategy model can be applied to these types of muscle contractions. Isometric contractions of the soleus muscle will be performed at two specified levels of force, which will be related to the maximum amount of force that a participant can produce. The force level will be maintained within a specified range of force. For example, participants may be asked to produce a force between 45% and 55% of their maximal force output. This would be defined as a 50% force level with a 10% range.

The dual-strategy hypothesis suggests that different

strategies should produce different neural excitation pulses. This pulse, proposed to modulate spinal excitability, is thought to be reflected by changes in the H-reflex (Gottlieb, Corcos, & Agarwall, 1989a). The H-reflex is a noninvasive test that can be used to measure changes in spinal cord excitability. Evaluated will be the neural mechanism (excitation pulse) that may govern the different strategies.

Independent variables for this study will include 1) target variables of 5% and 15% bandwidths and 2) target variables requiring two specific isometric force contraction levels (25% and 50%) which will be based on a participant's maximum voluntary contraction. Additional dependent






21



variables for evaluation of the neural excitation will include amplitude of the H-reflex, amplitude of the maximum M-response, and duration of H-reflex facilitation.



Research Hypotheses



The research hypotheses examined in two studies are as follows:



Study 1

The dual-strategy hypothesis was evaluated to determine whether this model can be extended to account for singlejoint isometric movements. It was hypothesized that, isometric movements can be classified according to either an SS or SI strategy.



1. Hypotheses tested for the SS strategy are as follows:

a. Single-joint isometric movements performed under the SS strategy will demonstrate that the

slope of the first time derivative of force (dF/dt)

will positively rise along different trajectories.

b. Q., and Qa, of agonist EMG activity will

positively rise with increases in bandwidths.

C. The 5% bandwidth will result in a smaller

amount of agonist EMG activity, slope of dF/dt, and






22



peak dF/dt when compared with the 15% bandwidth for

the respective force level.



With the SS strategy, peak accelerating torque, Q3c,f and Qac, have been shown to increase linearly with an increase in movement speed (Corcos, Gottlieb, & Agarwall, 1989b). For example, as the speed of the movement increases, there is a corresponding increase between peak accelerating torque and Q30 and Qac=* Therefore, both Q30 and Qa,, are sensitive to the speed of movement.


2. Hypotheses tested under the SI strategy are as follows:

a. Single-joint isometric movements performed with

the SI strategy will show that the slope of dF/dt

will rise along the same trajectory.

b. Q3. will be invariant with either increases or

decreases in maximum voluntary contractions.

c. The slope of force and Q.,, will result in a

positive relationship with an increase in the

maximum voluntary contraction.

d. A 50% maximum voluntary contraction will result

in a larger amount of agonist EMG activity, and

peak dF/dt when compared with the respective

bandwidth at the 25% force level.






23



For the SI strategy, the two phases of EMG activity (Q30 and Qac) have been represented by different relationships with peak accelerating torque. The variable,, Q30,, has been shown to be insensitive to the distance moved or inertial load. That is, Q3,) does not appreciably change with the manipulation of movement distance or load (Gottlieb, Corcos, & Agarwall, 1989b). In contrast, a linear relationship has been obtained'

between Qa= and changes in the independent variables (Gottlieb, Corcos, & Agarwall, 1989b).



Study 2



Different movement strategies were proposed to produce

different neural excitation pulses. This neural pulse should be reflected by changes in motor neuron reflex excitability. The hypotheses tested were that single-joint isometric movements performed under the SS and SI strategies produce different intensities and patterns of motor neuron reflex excitability prior to EMG activity onset.



1. Hypotheses tested for the SS strategy are when force

level is held constant and comparisons are made

across bandwidths,

a. The 5% bandwidth will produce the same duration

of premovement H-reflex facilitation when compared

to the 15% bandwidth.






24



b. The 5% bandwidth will produce a smaller slope

of premovement facilitation when compared to the

15% bandwidth.

c. The 5% bandwidth will produce a smaller Hreflex peak to peak amplitude during premovement

facilitation.



Because speed is the dominate feature of ballistic and step tasks, these movements would be classified as SS strategies according to the dual-strategy hypothesis and should exhibit pulse-height modulation. Numerous studies have shown that the onset of spinal facilitation occurs between 55 and 80 mns prior to EMG onset in ballistic plantar flexion tasks (Brunt & Robichaud, 1996; Eichenberger & Ruegg, 1983,1984; Manning & Hammond, 1990). Kagamihara et al. (1990) evaluated spinal facilitation in step and ramp plantar flexion tasks. The step task shows an earlier onset of spinal facilitation (55 ms prior to EMG onset) when compared to the ramp task (92 ms prior to EMG onset). These results appear to support the pulse-height control for SS strategies.



2. Hypotheses tested for the SI strategy are, when

bandwidth is held constant and comparisons are made

across force levels,






25




a. The 25% force level will produce a shorter

duration of premovement facilitation when compared

to the 50% force level.

b. The 25% force level will produce a smaller

slope of premovement facilitation when compared to

the 50% force level.

c. The 25% force level will produce a smaller Href lex peak to peak amplitude during premovement

facilitation.



The dual-strategy hypothesis suggests that the SI

strategy should reflect changes in the timing (pulse width) of spinal cord facilitation. Kagamihara et al. (1990) evaluated the time course of spinal facilitation in a ramp plantar flexion task. In the ramp task, the participant moved at a preselected cursor speed. The onset of spinal facilitation began at 92 ms prior to EMG onset. Additionally, Frank (1986) observed that there was an increase in the time period of spinal cord facilitation (70 Mns prior to ENG onset) with an anticipation timing task. The dual-strategy hypothesis proposes that the excitation pulse in an anticipation timing or ramp tasks should be classified as a SI strategy, since there is no requirement to control speed for either of these tasks. These examples appear to support pulse-width modulation of spinal cord facilitation,





26




since the onset of spinal facilitation occurs later than 50 ms prior to EMG activity onset.



operational Definitions



Terms defined for the purposes of this investigation are as follows:

Alpha motor neuron is a motor neuron that innervates extrafusal muscle fibers.

H/M ratio represents the percentage of motor neuron pool that is active at any particular time. This ratio is between the amplitude of the maximum M-response and the amplitude of the tested H-reflex.

H-reflex (Hoffman reflex) is an electrically stimulated monosynaptic reflex that excites the muscle spindle's Ia afferents. Action potentials are transmitted to the spinal cord where monosynaptic connections cause motor neurons to reach threshold, thereby causing the extrafusal muscle to fire.

Instructional variables are premovement instructions

that are given to the participant by the experimenter. This may include instructions to perform the task within a certain time frame or as accurately as possible. These are considered independent variables.






27



Isometric contraction is a muscular contraction where there is no external movement of the contracting limb.

M-response is an electrically stimulated direct motor

response. The maximum M-response that represents 100% of the activity of the specified muscle's motor neuron pool.

Measured variables are dependent variables that may include movement time, velocity, or ENG duration or amplitude.

Motor neuron reflex Pool excitability is the net sum of all facilitatory and inhibitory influences from afferent and descending influences on the alpha and gamma motor neurons.

Premotor time is the time from the visual signal to onset of ENG activity.

Reaction time is the time from the visual signal until voluntary movement is initiated.

Stratecrv is a set of consistent rules for muscle

activation that determine the pattern of EMG activity and torque trajectories.

Speed-insensitive strategy (SI) consists of movements that are insensitive to the speed of movement.

Speed-sensitive strategyv (SS) consists of movements that are sensitive to the speed of movement.

Spinal cord facilitation is an increase in the

excitability of the motor neuron pool that occurs prior to onset of EMG activity.






28



Basic Assumi3tions



The following assumptions were made:



1. Single-joint isometric movements can be classified according to different strategies. These strategies occur with changes in either specified force output or bandwidth. For example, when the task is to move to the same bandwidth at different levels of force output, the SI strategy is utilized.

2. Participants must remain on task. To help keep the participant on task, they were reminded to be as fast and accurate as possible after every 20 trials. Additionally, to help eliminate anticipation of the response, the muscle activity was monitored prior to the initiation signal. If muscle activity was present, participants were asked to relax their leg. Catch trials were also included to help the participant remain on task.

3. Differences between dependent variables in singlejoint movements are due to participants using either the SS or SI strategy.

4. The SS and SI strategy are represented by differences in the excitation pulse (i.e., pulse width vs. pulse height).

5. The excitation pulse is represented by changes in alpha motor neuron excitability that can be assessed by changes in the H-reflex.






29



6. The H-reflex is a measure of alpha motor neuron excitability.



Limitations



Limitations for this study are perceived to be as follows:



1. The participant may not use the appropriate strategy while performing the different tasks. In order to reduce this possibility, participants were verbally reminded after every block of 20 trials to produce the task as fast and accurately as possible

2. The proposed neural excitation pulse may not be

reflected by changes in the excitability of the motor neuron pool. This possibility was evaluated through pilot tests which indicated the existence of different neural pulses for the SS and SI strategy.

3. Isometric movements may not follow the dual-strategy paradigm. To help eliminate this possibility, pilot tests were run which indicated that isometric movements could be classified according to the dual-strategy paradigm.

4. Extraneous factors, such as any unnecessary

movement, caffeine, learning, or ankle movement may influence the level of spinal excitability. These problems were controlled by having participants: a) seated in a custom






30



chair that limits head and ankle movement, b) refrain from caffeine intake 12 hours prior to the test, and c) go though a series of learning trials prior to the experiment. Participants performed 50 learning trials for each force level and bandwidth. After completing the trials, participants continued to perform trials until they were accurate on 9 out of 10 trials.



Significance of the Study



movement dysfunction in patient populations, such as

Parkinson disease or stroke, is not well understood. Until research can identify the precise nature of dysfunction that produces cognitive, programming, or execution deficits, the most appropriate method for motor relearning or rehabilitation will be difficult to determine.

No comprehensive model analyzing programming and execution components of movement dysfunction has been established. A recent model of single-joint control, the dual-strategy hypothesis,, was developed by Gottlieb, Corcos, and Agarwall (1989). This model divides movements into speed sensitive (SS) and speed insensitive (SI) strategies, with the same initial impulse modulating both force trajectories and EMG activity. Specific relationships between these parameters differentiate the movement strategies. The "neural pulse" is indicated to be the initial impulse, which






31



is assumed to reflect motor neuron pool excitability. Different movement strategies should produce different neural pulse patterns. For example, the SS strategy should modulate neural pulse height, while the SI strategy may alter the neural pulse width.

Although Gottlieb and colleagues (1989) outlined specific relationships between torque trajectories and agonist activity with movement strategy, there was no attempt to examine the existence of the neural pulse or to show relationships between this pulse and motor programming. Evidence suggests different neural pulses are represented by changes in motor neuron excitability. For example, in an anticipation timing task that could be classified as an SI strategy, Frank (1986) ascertained an increase in the time period of motor neuron excitability prior to onset of EMG activity. This would be classified as pulse-width modulation. The neural pulse of different movements was evaluated by measuring changes in spinal excitability prior to movement onset. The dual-strategy hypothesis allows detailed evaluation of single-joint isotonic movement. Ambiguity existed as to whether isometric contractions can be classified by this model. The generalizability of this model was analyzed by evaluating programming and neural changes of single-joint isometric movements.

The evaluation of the parameters of spinal cord

excitability may aid in understanding the preparation of






32



motor neuron excitability prior to the movement. Additionally, a comprehensive model of single-joint control would enhance the ability of the practitioner to equate differences in force trajectories and parameters of muscle activity with specific neurological deficits. This information would assist in determining which neurological conditions produce cognitive, programming, or execution deficits in the performance of the desired movement. This may add information into the differences between force and timing components of movements, which would help in the analysis of movement dysfunction. For example, such information may help determine if Parkinson disease produces timing or force production (as suggested from the results of Horak, 1984) movement deficits. Physical therapists would benefit from this knowledge in designing patient treatment plans by being able to focus treatments on actual movement deficits. For example, therapists would be able to focus therapeutic interventions that help the patient either access or execute the motor program for the desired movement.














CHAPTER 2
REVIEW OF LITERATURE



Woodworth (1899) proposed that single-joint movements

were controlled by two phases of movement. Later, according to Latash, Bernstein described voluntary movements as being controlled by higher centers within the central nervous system (Latash, 1989). Higher centers were described as a "black box", with input/output relationships between internal structures controlling voluntary movement. In this analogy, there were many levels of control with each level having functional importance in the control of voluntary movements (Latash, 1989). Bernstein (1967) further evaluated the ability of the motor system to overcome the apparent redundancy of joint angles or "degrees of freedom" when performing a voluntary task. In particular, he noted that when blacksmiths used a hammer to hit a chisel, there was a small variability at the endpoint of movement, eventhough, there was a high variability of joint angles. From this observation, Bernstein deduced that the central nervous system controls movements by reducing or eliminating redundant degrees of freedom. At the time of Woodworth's and Bernstein's original hypotheses, the underlying mechanisms and structures within the central nervous system were not well understood.


33






34



No current model of multi- or single-joint control has been developed to fully analyze programming and executive components of single-joint movements. Research oriented toward the evaluation of single-joint movements may be advantageous because of the inherent difficulties with attempting to control multi-joint movements. However, a recent model of single-joint movements by Gottlieb, Corcos, and Agarwall (1989a) has described characteristic relationships between EMG activity and torque trajectories during isotonic movements. This model, the dual-strategy hypothesis, divided movements into SS and SI strategies, with the different strategies utilizing a different set of rules for single-joint isotonic tasks (Gottlieb, Chen, & Corcos, 1995; Gottlieb, Corcos, & Agarwall, 1989a). However, ambiguity existed as to the precise nature of these relationships in isometric tasks (Corcos, Agarwall, Flaherty, & Gottlieb, 1990; Gottlieb, Corcos, & Agarwall, 1989a).

The utilization of the SS or SI strategy was proposed to correspond with a different modulation of excitation pulse converging onto the motor neuron pool (Gordon & Ghez, 1987b; Gottlieb, Corcos, & Agarwall, 1989a; Jaric, Corcos, Agarwall, & Gottlieb, 1993). The excitation pulse should be reflected by different patterns of motor neuron pool excitability prior to the movement. For example, different strategies may modulate either the amplitude or duration of the excitation pulse. The existence of the hypothesized excitation pulse






35



has not been investigated. However, the neuromuscular system does produce characteristic changes in motor neuron pooi excitability prior to a voluntary movement (Brunt & Robichaud, 1996; Eichenberger & Ruegg, 1984; Riedo & Ruegg, 1988). These changes have been investigated in conjunction with premotor reaction, response, motor, and movement times. Simultaneous investigation of motor neuron pool excitability during both an SS and SI voluntary isometric movement may allow the analysis of the hypothesized excitation pulse. That is, a comparison of spinal excitability patterns could be evaluated between the SI and SS strategies.

This review will examine the relationships between

torque trajectories and agonist ENG activity of single-joint isometric movements to evaluate whether the dual-strategy hypothesis can be extended to account for isometric movements. The review of single-joint movements will cover both motor programs and models of programmed movements. The discussion will be limited to models of motor control that are testable in terms of central nervous system functioning. This is not to discount other models, such as the dynamical theory or Kelso's co-ordinateive structure model. However such models do not specify the control of movement in terms of a structured code within the central nervous system and therefore are not currently testable using a neurophysiological approach to movement. The specific models that will be discussed include: Fitts' speed/accuracy






36



tradeoff model, the impulse timing model, the pulse height model, and the dual-strategy model. Reference to the dualstrategy model will include discussion of the SS and SI strategies during both isotonic and isometric contractions. The relationship between single-joint movements and EMG activity will be analyzed. Additionally, the proposed excitation pulse which may modulate the dual-strategy hypothesis will be reviewed. This analysis will include the relationship of motor neuron excitability with 1) voluntary movement, 2) force output, 3) reaction time, and 4) the SS and SI strategies.



Motor Programs


Lashely (1951) was one of the first motor theorist to propose that movements were centrally controlled. The central control of voluntary movements has subsequently been described as the "memory drum theory" by Henry and Rodgers (1960) and more recently as a "motor program" by Keele (1969) and Schmidt (1976). In the memory drum theory, Henry and Rodgers (1960) ascertained that reaction time increased with the number of required movements. It was proposed that movement information was loaded into a buffer, termed a "memory drum", after the initiation signal. The larger the number of required movements subsequently increased buffer loading time and reaction time. This interpretation has been supported by the work of Sternberg and colleagues (1978).






37



Schmidt (1976) expanded on the motor drum theory by

introducing the concept of a motor program that controlled voluntary movements. A motor program was defined as a centrally stored restructured set of muscle commands which carry out movement without feedback about the goal of the movement. However, there were two areas of concern with the concept of a motor program (Schmidt, 1975). The first difficulty encompassed a storage problem, where the assumption was a one-to-one relationship between a motor program and each motor response. This assumption required a large number of motor programs and central storage area to account for all possible movements. The capability to store numerous motor programs have not been disproved, but the question of storage remained an issue. The second concern involved the "novelty problem", described as the ability to produce novel movements under new circumstances and environmental conditions which require new, adapted, or modified movement for success. The problem was whether a new movement was adapted from a previously stored motor program or required the generation of an entirely new motor program.

The "generalized motor program" or "schema theory" was

therefore formulated to address the concerns with storage and novel movements (Schmidt 1975). The schema theory, like the original motor program, acknowledged a centrally stored, restructured set of muscle commands for carrying out a class or category of movements. However, the assumption of a one-






38



to-one relationship between the motor program and a motor response produced was replaced by a one-to-many view (Schmidt, 1975). Variations of motor responses within a movement class arose from altering parameters of the motor program (or schema), such as movement speed or force. This perspective provided an explanation for the generation of novel movements by changing the response specifications of the schema for a class of movements.

A schema was further theorized by Schmidt (1975, 1976) to be composed of invariant and variant features. Invariant features were aspects which were fundamentally structured in the schema and common to the generalized motor program. The order of events, temporal structure of muscle contractions, and the relative force have all been implicated as invariant features of the schema. Variant features are those dimensions or parameters of the schema that were relatively superficial and temporary. Selecting variant features or parameters such as force and duration afforded a variety of movement responses within a generalized motor program (Brooks, 1986; Schmidt, 1988).

Investigations about the production of a motor program have addressed issues such as the required programming time related to movement complexity (Fischman, 1984), programming force, timing components, (Ivry, 1986), and the effects of force variation on the motor program (Ito, 1990). The effects of the above parameters on motor programming, have






39



been evaluated during simple and choice reaction time studies by fractionating reaction time into premotor and motor time. Premotor time has been defined as the time between the onset of a stimulus for movement and the onset of EMG activity. Premotor time represents the time required to centrally organize, select and transfer the appropriate commands to the muscles for initiation and execution of the movement. Motor time can be described as the interval of time between the onset of EMG activity in the selected muscle above the baseline activity and the initiation of the response. This may be representative of the peripheral processes of the response. Therefore, reaction time should equal to the sum of premotor time and motor time (Schmidt, 1988).

Fischman (1984) has stated that premotor time served as a more valid representation of programming time rather than reaction time alone. He asserted, that without fractionation of simple reaction time, it would be difficult to discern whether changes in total reaction time were due to central processes, peripheral processes, or another variable. Central processing time was identified as the time required to access the underlying motor program. Based upon this assumption, both Fischman (1984) and Ito (1990) used fractionated reaction time in their methodology to distinguish difference between the central and peripheral processes presumably associated with motor programming. The regularities of single and multi-joint movements have led






40




investigators to propose various models for performing voluntary movements.



Models of Programmed Movements


Single-joint movements have been described by consistent relationships between observable features such a EMG activity (agonist duration and magnitude) and kinematic trajectories (force-time curves). For example, the duration of agonist EMG burst activity can be associated with the acceleration of the movement. That is, a positive linear relationship existed between the duration of agonist ENG activity and movement acceleration. This, plus other, observations have led to the view that a general relationship exists between agonist EMG activity and kinematic measures (Cooke, & Brown, 1994). However, the above assumption was contrasted by Hoffman and Stick (1993) and Gottlieb et al. (1989). These researchers proposed that agonist burst EMG activity and kinematic parameters can be independently specified by the nervous system. Gottlieb et al. (1989) went further to describe all single-joint movements as being based upon two separate movement strategies. A strategy "consists of sets of rules for muscle activation {that lead tol patterns of muscle torques and EMG activity ... [which] are highly and consistently correlated irrespective of task" (Gottlieb, Chen, & Corcos, 1995). Therefore, single-joint movements may






41



be classified by different strategies which are based upon rules for EMG and kinematic measures.

Rules for single-joint movements consist of consistent relationships between two parameters of movement. Rules may exist among both dependent and independent measures. Dependent measures may include movement time, velocity, EMG amplitude, EMG burst duration, or errors. An example of rules among dependent variables includes the relationship between peak agonist EMG activity and peak velocity, which can both be associated with distance (independent variable) moved. However, there may be an independent increase or decrease among the dependent variables which can change with the independent variable. Independent variables are under control of the experimenter and include the manipulation of target size, movement distance, movement time or inertial load (Gottlieb, Corcos, &-Agarwall, 1989a). Independent variables were considered to be directional and may affect the dependent variables, however, dependent variables could not affect the independent variable. Rules existing between independent and dependent variables have also been identified (Gottlieb, Corcos, & Agarwall, 1989a). This can be seen in relationships between movements to various distances, bandwidths, and with different inertial loads. For example, as target distance increased (independent variable) both movement time and velocity increased (dependent variables). Additionally, an increased target size (independent variable)






42




resulted in a decrease in movement time with a corresponding increase in velocity (dependent variable) (Fitts, 1954; Gottlieb, Chen, & Corcos, 1995; Gottlieb, Corcos, & Agarwall, 1989a).

Based upon the experimental design, a previous dependent variable may become a independent variable. For example, if movement time was constrained it will become an independent variable. These fluctuating relationships among dependent and independent variables have confounded the issue of the relationships between these variables of single-joint movements. This tends to limit the generalizability of most models in explaining many aspects of single-joint movements. Eventhough controversy exists regarding an appropriate model for a model for single-joint movements, evidence suggests that these movements will be programmed in advance (Alexander, & Crutcher 1990a, 1990b; Crutcher & Alexander 1990).

Voluntary ballistic movements are planned and programmed in advance (Brooks, 1986). An analysis of Woodworth's (1899) original concept of control of limb trajectory provides insight into programmed movements. Woodworth (1899) theorized movements are controlled by two phases of movement. The first phase determined the lim~b's initial trajectory while the second phase directed the limb to the final target. The initial phase may be considered the motor program for implementing the movement, therefore there should be some






43



predictable patterns that are consistent from task to task. These predictable patterned responses have eluded motor control theorists. The second phase was postulated to be influenced by visual or kinesthetic input which brought about movement corrections which may not be preprogrammed (Rosenbaum, 1991; Schmidt, 1988).

The notion that ballistic movements are totally

preprogrammed arises from the concept that rapid movements are to fast to contain corrective adjustments (Desmendt & Godaux, 1978). However, this point has been disputed because corrective adjustments can be made in a time period too brief to be based purely on feedback from the movement (Gordon & Ghez, 1989a; Vicario & Ghez, 1984). The implication is that movements may be adjusted by a feed forward mechanism. Even though certain aspects of the movement may be modified, the earliest portion of the movement may still be preprogrammed.

No comprehensive model exists in which programming and execution components of single-joint movements have been analyzed. Woodworth's (1899) original hypothesis has not has not been discounted and many subsequent investigators have suggested various models for governing single-joint movement. These include the models proposed by Schmidt and colleagues' (1979) impulse-timing model, Fitts' (1954) speed-accuracy model, Gordon and Ghez's (1989a) pulse-height model, and Gottlieb and colleagues, (1989a) dual-strategy model. These models all incorporate different descriptions between kinetic






44



and kinematic variables, relationships between EMG and trajectory control, or relationships between movement speed and accuracy. However, most models appear to explain only certain aspects of the movement and can usually only be associated with specific experimental paradigms.



Fitts' model


In 1954, Fitts evaluated the relationship between speed and accuracy of simple aiming movements. Fitts assessed a reciprocal tapping task that was not thought to require central processing, therefore fundamentally different from the simple aiming movements used by Woodworth (1899). In Fitts' paradigm, subjects were instructed to tap as fast and accurately as possible between two targets with the manipulation of target width and inter-target distance (amplitude). Relationships between movement time, movement amplitude, and target width were established as a result of this paradigm. The relationship was related to the difficulty of the task.

Task difficulty, known as the index of difficulty (Id), was ascertained to be the mathematical relationship of two times the logarithm (to the base 2) of the movement amplitude divided by the target width. That is, Id = 1092 2 A/W, where A equals the amplitude of the movement and W equals the width of the target. This indicated that movement time increased






45



when tapping movements were directed toward smaller targets or targets were moved further away (Fitts, 1954). The movement time to distance relationship was presumed to be produced by corrective submovements which were executed after visual feedback. This has become known as the speed/accuracy trade-off which relates to the accuracy component of the task (Newell & Corcos, 1993; Schmidt, Zelaznik, Hawkins, Frank, & Quinn, 1979). That is, depending upon whether a person performs the task "as fast" or "as accurately" as possible, the other variable will presumably be sacrificed.

Schmidt (1979) and Meyer et al. (1982) analyzed

isometric movement under the assumption that movements were not comprised of a set of discrete submovements. Under this assumption, a linear relationship between speed and accuracy resulted which were different from the logarithmic relationship produced when distance was controlled. This was contrary to Fitts' law and Meyer (1990) suggested movements that do not utilize feedback-based corrections will not exhibit the logarithmic relationship between speed and accuracy (Meyer, Abrams, Kornblum, Wright, & Smith, 1988; Meyer, Smith, & Wright, 1982). There have been several research reports which support the linear speed/accuracy trade of f in isometric movements (Wright & Meyer, 1983; Zelaznik, Mone, McCabe, & Thaman, 1988; Zelaznik, Shapiro, & McColsky, 1981). Additionally, Klapp (1975) showed that with easy movements (reaction times less than 200 ins) there was no






46



relationship between the index of difficulty and movement duration. Keele (1968) also established a linear relation between movement speed and spatial variability.

Crossman and Goodeve (1983) proposed a theory to account for Fitts' law. In this theory, based upon feedback control of movement, it was assumed that a series of corrections can impact response accuracy. Movements made over longer distances or to smaller targets presumably required more corrections and therefore more time. Keele (1986) also showed this feedback-based assumption for the control of movements which appeared to support Fitts' law. Keele (1986) indicated that movement time was related to the number of corrections required to achieve the target. This notion accounted well for Fitts' law and supported the feedbackbased idea of motor control. However, the use of feedback may pose a problem with Fitts' law.

Evidence appears to indicate that subjects have

difficulty processing feedback during movements that take less than 200 ms. (Schmidt, 1976). Therefore, fast reaction times may not be conducive to on-line response corrections. From this perspective, the feedback-based idea for Fitts' law can not account for all the experimental data. Crossman and Goodeve (1983) evaluated the limitations of error correction and concluded that short movement sequences may be run off without any peripheral feedback. This notion appeared to support the assumption that movements may be performed by






47



restructured muscle commands, which have been termed a motor program or schema (Keele, 1969; Schmidt, 1975, 1976).

The Fitts' paradigm involved a closed loop process which required sufficient time for feedback. This was contrary to Woodworth's (1899) assumption that rapid movements do not require visual control. Behavioral evidence appeared to support Fitts' law for single arm movements performed in both air or water (Fitts & Peterson 1964; Kerr, 1973). More recent models of motor control have extended past the observational data, which Fitts' model was based upon. These models, which include the impulse timing model, pulse height model, and the dual-strategy model, all have added a central nervous system component when explaining the control of voluntary movements.



Impulse Timing Model


The impulse timing model was developed using a centrally mediated impulse which controlled the movement. The impulse of electrical activity delivered to muscles may be characterized by acceleration and force-time functions with a motor program controlling the onset, amplitude, and duration of the impulse (Schmidt, 1976). The total impulse should be proportional to the area under the acceleration-time curve, with impulse duration being proportional to movement time,






48



while impulse height should be inversely proportional to movement time (Jeannord, 1991; Schmidt, 1976).

The impulse timing model include the following

force/time assumptions: 1) movements that are velocityconstrained produce a sinusoidal like force-time function, 2) force-time functions are controlled as a unit, such that with alteration of muscular force the force-time function will be proportionally recalled, and 3) force-time functions can be captured by the impulse's amplitude and duration (Jeannord, 1991; Schmidt, 1988; Schmidt et al., 1979). The impulses' amplitude and duration may be controlled by altering either the force burst or the time over the transmitted force (Schmidt, 1988). Therefore, force-time and the duration of force can vary independently and be appropriately recalled to the other parameter without affecting the goal. For example, force may be increased while holding the timing of the movement constant. This would produce movements with the same duration, however the amplitude of the movement would became progressively larger with increased force.

The applied impulse may relate to the magnitude of force output and may dictate the pattern of movement and muscle contraction relative to the desired force. There appears to be a proportional relationship between the applied impulse and force and force variability, with the variability of the impulse being related to the impulse duration. One problem with this model was the implication that these impulses were






49



only sent to agonist muscles. Therefore, no account for deceleration forces was provided (Bizzi, 1982, 1984; Meyer, 1982).

The assumption that the force-time function can be

rescaled has been evaluated. Meyer (1982) showed that time and force were rescalable functions with the intensity and duration of the impulse being similarly scaled to the amplitude and period of a sine function. An implication of the rescalable properties of force and time implied that both these parameters can vary independently. For example, Freund and Budinger (1978) showed that force would be rescaled when movement time remained constant in an aimed target movement. Additionally, movement time and distance have been shown to change by altering the entire acceleration/time profile as a unit (Schmidt, 1988).

Behavioral evidence does not support the force-time rescalable function. Zelazink et al. (1986) analyzed kinematic data of aimed hand movements and showed a strong linear relation between target width and distance/time. However, time to peak acceleration was not related to time, but was fixed across different values of time. The impulsetiming model utilizes kinematic as a evidence for a central mediated impulse which controls movement. However, this model does not include any explanation as to the origin of the hypothesized impulse, nor does it incorporate any measures or muscle activity into the model.






50



Pulse-Height Model


Gordon and Ghez's (1987a) pulse-height model expanded on the previous models by adding EMG activity and an instructional variable as a parameter which can influence single-joint movement. Instructional variables were premovement instructions that were given to the subject by the experimenter. For example, the subject may have been told to execute the task within a certain time frame, as fast as possible, or as accurately as possible.

Gordon and Ghez (1987a) evaluated isometric forces to establish the presence of a control policy that modulated force production. Subjects performed elbow flexion isometric movements under the instructional sets of "fast" or "accurate". They were requested to make no corrective adjustments during the movement. Through a series of experiments, invariant relationships between limb trajectories andEMG activity allowed the inference of control policies which appeared to be operated by a motor program.

The tasks were designed to have subjects perform fast and accurate movements under three different force (37% to 53% of a subject's maximum voluntary contraction at the ratio of 1:2 and 1:3) levels at a 16% bandwidth. Each condition revealed similar linking variables of the initial peaks of the 1st (Force velocity dF/dt) and 2nd (Force acceleration






51



- d2F/dt2) time derivative of force being strongly predictive of peak force. However, there were several differences noted between these conditions. The accurate condition was less variable and was related to longer force rise times. There was no dependence between force rise time and peak force, with force rise time remaining around a constant value. There was clear triphasic EMG activity with increased agonist activity in the fast condition. Agonist muscle activity increased with force, while the antagonist muscle activity and duration remained unchanged. However, the antagonist muscle activity varied with force rise time (Ghez & Gordon, 1987; Gordon & Ghez, 1987a).

The pulse-height model suggested that torque pulses were generated by rules governing pulse-height control, with the first measurable feature of the response predicted the entire response. That is, torque pulses were proportionally scaled to allow responses of different amplitudes. The scaling of this impulse can be influenced by the instructions provided to the participant. The instruction to move accurately affected the rate of rise of force, while force rise time was held constant. In the fast condition, force rise time was not regulated (Gordon & Ghez, 1987a). However, there was no attempt to utilize this model to explain results of past studies or to provide an origin for the actual control of the torque pulse.






52




Dual-Stratectv Model


Gottlieb, Corcos, and Agarwall (1989a) developed the

dual-strategy model as a universal model which described the control of single-joint voluntary movements. This model defined the programed aspects of movement and accounted for relationships among dependent and independent variables. Single-joint voluntary movements across different distances, with different inertial loads, toward targets of different widths and over a wide range of experimentally manipulated velocities can be explained using the dual-strategy model (Latash & Gottlieb, 1992). This model was based on three propositions: 1) movements were planned according to "strategies" which were based on a set of rules associated with the nature of the task and the instructions provided to the performer, 2) the control of single-joint voluntary movements was based on at least two strategies, which include the SI and SS strategies, and 3) the choice between the SI and SS strategies depends on whether or not movement speed was controlled during the task (Gottlieb, Chen, & Corcos, 1995; Gottlieb, Corcos, & Agarwall, 1989a).

The dual-strategy model was based on regulation of the net presynaptic input on the alpha motor neurons of the agonist and antagonist muscles. The control signal was modeled as a rectangular excitation pulse, whose origin was not delineated; however, parameters which may influence the modulation of the excitation pulse were defined (Corcos,






53




Jaric, Agarwall, & Gottlieb, 1993; Gottlieb, Corcos, Agarwall, & Lattash, 1990). The excitation pulse's amplitude and duration was presumably controlled by the central nervous system. The excitation pulse was thought to be controlled by altering either the excitation pulse's height or width. That is, an increase in the height of the pulse would indicate pulse-height control, while an increase in the length of the pulse would correspond to pulse-width control. Therefore, two different mechanisms for controlling the excitation pulse which correspond with the two different strategies of movement were indicated. According to Gottlieb, Corcos, and Agarwall (1989a), this excitation pulse should be reflected by the changes in the status of the excitability of the spinal motor neuron pool (Gottlieb et al., 1990).

The dual-strategy model assumed that agonist ENG

activity and limb trajectories were consequences of the same initial excitation pulse. Relationships were ascertained between peak accelerating torque and EMG activity which were dependent upon whether or not movement speed was controlled during the task. To further analyze EMG activity, agonist activity was divided into two separate time periods. The first time period (Q30) consists of the first 30 mns of agonist activity. The agonist activity after the initial 30 mns until the end of acceleration was described as Q acceleration (QC,). Depending upon the movement strategy (SS or SI), different relationships were observed between peak accelerating torque






54




and the two phases of EMG activity during isotonic movements (Corcos, Gottlieb, & Agarwall, 1989; Gottlieb, Corcos, & Agarwall, 1989b). In addition to differences in EMG activity, there were other measures which differentiated the SS and SI strategies.



Speed-insensitive strategyv


The SI strategy was ascertained to be the standard and default strategy for single-joint voluntary movements. This strategy was considered to be insensitive to movement speed. For example, SI strategies were utilized when movements were performed to the same bandwidth but to different distances, or when the task was to move the same distance but different loads were applied to the limb. A linear relationship was obtained between peak accelerating torque and the measured variables of peak torque, peak velocity, and movement time when single-joint voluntary movements were performed under this strategy. These variables initially rose along the same trajectory and these relationships held when task variables such as distance moved or inertial load were changed. For example, as distance decreased, there was a linear decrease in peak velocity, peak torque, and movement time (Gottlieb, Corcos, & Agarwall, 1989a, 1989b).

The burst activity of the agonist muscle was shown to

initially rise at a uniform rate. The duration of the rising






55




phase and peak amplitude of the EMG activity varied in proportion to changes in distance or load. The initial acceleration segments over the first 30 to 50 mns were insensitive to the distance moved. There were different relationships with peak accelerating torque and the two phases of agonist EMG activity. The variable, Q.., was shown to be insensitive to the distance moved or inertial load (task variables). That is, Q3, did not appreciably change with the manipulation of movement distance or load (Gottlieb, Corcos, & Agarwall, 1989b; Jaric, Corcos, Agarwall, & Gottlieb, 1993). In contrast, there was a linear

relationship for Qa, with changes in the task variable (Gottlieb, Corcos, & Agarwall, 1989b). The antagonist bursts vary similarly and had latencies that increased with the magnitude of the task variable. Additionally, the SI strategy was implicated as the default strategy for movements where speed was not explicitly or implicitly implied as the independent variable (Gottlieb et al., 1990).



speed-sensitive strategy


The SS strategy consisted of movements where speed was either explicitly or implicitly implied as a independent variable. For example, subjects were instructed to move in a given time period, to different sized targets at the same distance or to move as rapidly as possible. Peak





56




accelerating torques and peak velocity were shown to linearly increase with an increase in movement speed when movements were performed under this strategy (Jaric, Corcos, Agarwall, & Gottlieb, 1993). However, these variables initially rose along different trajectories. Peak accelerating torque and both Q3,0 and Q.. both increased linearly with increased movement speed. For example, as movement speed increased there was a corresponding increase between peak accelerating torque and Q,,,. This relationship was also established for Q30 (Corcos, Gottlieb, & Agarwall, 1989; Gottlieb, Chen, & Corcos, 1995). Therefore, both Q30 and Qac, were sensitive to movement speed. Antagonist bursts varied similarly and had latencies that increased with movement time. Additionally, the onset of antagonist EMG activity was delayed with an increase in movement time (Corcos, Gottlieb, & Agarwall, 1989). The ability to differentiate the SS and SI strategies was based upon specific pattern differences in agonist and antagonist EMG activity (Corcos, Jaric, Agarwal, & Gottlieb, 1993; Jaric, Corcos, Agarwal, & Gottlieb, 1993). These differences should be reflected by different build up patterns of motor neuron excitability prior to the movement.



Excitation Pulse


The dual-strategy model, based upon an excitation pulse, should account for the different patterns of EMG activity and






57



muscle torque trajectories which appeared to be dependent upon the movement strategy (Gottlieb, Corcos, & Agarwall, 1989a). This excitation pulse may be modified depending upon which movement strategy was utilized by a person (Gottlieb, Corcos, & Agarwall, 1989b; Newell & Corcos, 1993).

Single-joint movements performed under the SI strategy produce different movement profiles as a consequence of changing the width of the excitation pulse. The width of the excitation pulse should vary proportionally with changes in task variables, such as distance moved or inertial load. However, the intensity of the excitation pulse should be held constant. This has been described as pulse-width modulation (Gottlieb, Corcos, & Agarwall, 1989b).

single-joint movements performed under the SS strategy produce different movement profiles as a consequence of a change in the height of the excitation pulse, while the width of the excitation pulse should remain uniform. The uniform intensity of the excitation pulse should not be a product of movement speed, with submaximal speeds being associated with submaximal intensities. The submaximal intensities may be selected features of the pattern of movement control. The intensity of the excitation pulse should vary with changes in movement speed. For example, movements performed to the same distance at different bandwidths should show modulation of the intensity of the excitation pulse, while the duration of the excitation pulse should be held constant (Gottlieb,






58




Corcos, & Agarwall, 1989b). This has been described as pulse-height modulation (Gottlieb, Corcos, & Agarwall, 1989b). The dual-strategy model was initially utilized to explain single-joint isotonic movements, however evidence suggests this model may be applied to isometric movements.



Isometric Contractions


Corcos, Agarwall, Flaherty, and Gottlieb (1990) extended the dual-strategy model of single-joint control to include isometric contractions. They proposed that isometric contractions were controlled in the same manner as singlejoint isotonic movements, but this assumption has never fully been supported. Gordon and Ghez (1987a) appear to support this assumption because their evaluation of fast and accurate isometric contractions established that the first measurable features of a response (e.g. the first and second time derivatives of force) could be used to predict the entire response. In the evaluation of isometric contractions by Corcos et al. (1990), movements performed under the SS or SI strategies appeared to be a blending of the two strategies patterns.

An analysis on the research on isometric contractions by Corcos et al. (1990), seemed to indicate some limitations in their research design and subsequent analysis of their data. Gottlieb, Corcos, and Agarwall (1989b) previously stated that






59




peak accelerating torque was the linking variable that could explain correlations between dependent and independent variables. The research analyzing isotonic movements demonstrated specific relationships between peak accelerating

torque and EMG activity (Q30 and Qac) with the manipulation of independent variables, which could differentiate the SS and SI strategies (Corcos, Gottlieb, & Agarwall, 1989; Gottlieb, Corcos, & Agarwall, 1989b). However, under the isometric condition there was insufficient analysis between torque variables and the dependent variables to draw sufficient conclusions.

Additionally, there appeared to be some conflict in the classification of the SI and SS tasks. For example, subjects performed four different levels of a maximum voluntary contraction to the same bandwidth under two different instructional sets. The instructions were to perform the task as fast and accurately as possible and then at a comfortable speed. The performance of a task to the same bandwidth at four different levels of a maximum voluntary contraction would be classified as an SI strategy. However, when the instructions to move at a comfortable speed were added, the task became a combination of SS and SI strategies. There appeared to have been some interplay between the SS and SI strategies in this isometric study.

Researchers concerned with isotonic single-joint

movements have analyzed torque trajectories and EMG activity






60



changes between the same amount of force and bandwidth changes of 3%, 6%, 9% and 12% of a specified maximum voluntary contraction. This task should be classified as an SS strategy. However, this paradigm appeared to indicate a division between the SS and SI (Corcos, Agarwal, Flaherty, & Gottlieb, 1990). Bandwidths were manipulated between 1.50 to

120 for a 540 target in a study by Corcos et al. (1990). A visual analysis of the data appeared to indicate that relationships between peak accelerating torque and EMG activity were dependent upon the specific bandwidth. That is, the relationship between peak accelerating torque and Q,"

seemed to be invariant with the bandwidths of 1.5* and 3* for a 540 target, while there was an apparent linear relationship between these variables and the larger bandwidth targets (60

and 120 of a 541 target). For example, when subjects performed under the larger bandwidths they appeared to be performing under an SI strategy strategies (Corcos, Gottlieb, & Agarwal, 1989). That is, the bandwidth became large enough to no longer influence the movement. This appeared to suggest that for smaller targets individuals performed under the SS strategy, while performance under the SI strategy occurred to the larger targets. It appeared there may be a critical bandwidth whereby subjects performed under either the SI or SS strategy. Further research is required to differentially determine if isometric contractions operate under a dual-strategy hypothesis. Additionally, there






61



appears to be characteristic muscle activity patterns which differentiate the SS and SI strategies.



Electromyographic Activity


Ballistic movements have been described by

characteristic patterns of EMG activity where an agonist burst was followed by a silent period prior to a second agonist burst (Halite & Khoshibin, 1980). EMG activity demonstrated a temporal pattern between the agonist and antagonist muscles. There was a discharge from neural cells in the motor cortex prior to and during voluntary movement (Evarts & Fromn, 1978; Freud & Budinger, 1978; Garland, 1972). These observations appear to provide evidence that the early stages of ballistic movements can be centrally programmed and under the control of higher centers (Hallett & Khoshibin, 1980; Nagasaki, Irie, & Nakamura, 1983).

Several relationships have been established between

force generation and EMG activity. (Yonenda, Oishi, & Ishida, 1983) The timing and duration of agonist and antagonist muscle activity during force generation produce specific EMG patterns, which were dependent on velocity, external perturbations, inertial load, and accuracy of a voluntary movement (Gottlieb, Corcos, & Agarwall, 1989a; Meinch, 1984; Sanes & Jennings, 1984). The initial burst of EMG activity






62



has been shown to be unaffected by mechanical perturbations (Hallett & Khoshibin, 1980).

The amount of muscular activity dependents upon both the level of force and the rate of rise of force. Yondenda et al. (1983) evaluated the relationship between the amount of muscular activity and variations in force output. Two characteristics emerged between force production and muscle activity. The amount of muscular activity increased with both increased time to peak force and force output. When the time to peak force was under 150 ms, the amount of muscular activity increased with increasing force levels. However, when the time to peak force was above 150 ms, the amount of muscular activity increased with increasing times to peak force (Yonenda, Oishi, & Ishida, 1983).

The amount of EMG activity has been evaluated in

conjunction with the rate of force development over a range of joint angles (Nagasaki, Irie, & Nakamura, 1983). The rate of force development was constant during the first 70 ms over

joint ranges of 300 to 1200 of elbow flexion. This was evident even though the effective muscular force of elbow extension may be affected at different angles of movement. Changes in timing had no effect on the pattern over which the rate of muscle tension was developed. The amount of discharge during ballistic contractions varied more than during ramp contractions. Perhaps both speed and amount of force may contribute to EMG activity and force relationships.






63



The different discharge patterns of EMG activity determined by preprogramnming may be related to different movement strategies (Nagasaki, Irie, & Nakamura, 1983).

The amount of EMG activity can be influenced by target force. Suzuki, Yamazaki, and Matsunami (1994) evaluated power grip contractions to target forces between 16.7% and 100% of a maximum voluntary contraction (Suzuki, Yamazaki, & Matsunami, 1994). The rate of rise of force along with the amount of EMG activity increased with peak force, while the time to peak force was held constant when contractions were between 16.7% to 50% of a subject's maximum voluntary contraction (Suzuki, Yamazaki, & Matsunami, 1994). When the target forces were 66.7% or greater of a maximum voluntary contraction, the rate of rise of force rise and amount of EMG activity leveled off while the time to peak force was prolonged. However, after 66.7% of a maximum voluntary contraction there was no further increase in ENG activity. This corresponded with the amplitude of EMG activity being modulated up to 50% of a maximum voluntary contraction, while the amplitude of EMG activity leveled of f with the duration being modulated above 66.7% of a maximum voluntary contraction.

Contractions of 50% of a maximum voluntary contraction or less were suggested to support the pulse-height model, while contractions greater than 66.7% of a maximum voluntary contraction appeared to support a pulse- width/duration model






64



(Corcos et al., 1990). Additionally, the initial burst duration of EMG activity were identical for 16% through 100% of a maximum voluntary contraction (Suzuki, Yamazaki, & Matsunani, 1994).

motor unit recruitment may be indicative of changes in EMG activity and reflect motor programming. Solomonow, Baratta, Shoji, and D'Ambrosia (1990) noted a strong dependence of intramuscularly recorded EMG activity and force production. When force output increased up to 50% of a maximum voluntary contraction, there was a linear relation between intramuscular EMG activity and force output. After 50% of a maximum voluntary contraction the EMG-force curve became progressively non-linear. The data indicated a switch from a linear to a non-linear relationship between EMG activity with increasing force production.

Solomonow (1990) indicated a change in the muscle

control strategy can appear around 50% of a subject's maximum voluntary contraction. Additionally, the pattern of motor unit recruitment was shown to be different between muscles, which possibly related to different fiber types (Solomonow et al., 1990). That is, there was a more linear EMG-force curve for slow twitch muscles when compared to fast twitch muscles (Solomonow et al., 1990). Muscle type and the apparent switch in strategies which occurred around contractions that were greater than 50% of a maximum voluntary contraction may confound the evaluation of motor control models.






65




Additionally, motor neuron pool excitability which has been related to EMG activity, may provide further evaluation into the control of single-joint movements (Butler, Yue, & Darling, 1993).



Motor Neuron Pool Excitability


Control of single joint movements has been described by the dual-strategy hypothesis. This hypothesis proposed different independent variables dictated the different movement strategies as a consequence of changes in the excitation pulse. According to Gottlieb, Corcos, and Agarwall (1989a), this excitation pulse should reflect the status of the excitability of the spinal motor neuron pool.

Motor neuron pool excitability has been assessed by

evaluating changes in the Hoffman reflex (H-reflex) (Angel & Hoffman, 1963; Honore, Demaire, & Coquery, 1983; Schieppati, 1987). The H-reflex can been defined as an electrically stimulated monosynaptic reflex which primarily excites the muscle spindle's Ia afferents. Additionally, the overall reflex excitability may be influenced by excitation of Ib afferents from Golgi tendon organs, group II afferents from muscle spindles and larger cutaneous afferents (Schieppati, 1987). The impulses travel to the spinal cord where monosynaptic and pollysynaptic connections cause motor






66




neurons to fire, thereby causing the extrafusal muscle to contract (Honore et al., 1983; Schieppati, 1987).

The validity and reliability of the H-reflex (as a

measure of the excitability of the alpha motor neuron pool) has been evaluated by Crayton (1981) and Hugon (1973). Because of the variability of the H-reflex, Crayton (1981) and Hugon (1973) concluded that one test measurement should consist of the average of at least 5 to 7 (and up to 20) Hreflex measurements. Crayton (1981) further demonstrated a wide inter-subject variability in reflex recording; however, there was a small test/retest variability for each individual. Because of this variability, Crayton (1981) suggested that changes in the H-reflex should only be analyzed relative to each subject's baseline reflex (Hoffman, 1973).

Several methods can be used to evaluate changes in motor neuron reflex excitability. The change in excitability has typically been evaluated as a percentage change of the size of the test H-reflex in comparison to a control (baseline) reflex (Hugon, 1973; Schiepatti, 1987). This method of analysis requires an understanding of how changes in stimulus intensity can influence the H-reflex amplitude. As stimulation intensity increased, the amplitude of the Hreflex will follow a bell-shaped curve. That is, the Hreflex will become progressively larger up to a critical point, after which the amplitude gets progressively smaller






67




until the reflex fades away. Therefore, test H-reflex alterations elicited by a given conditioning stimulus depend not only on the conditioning volley, but also on the amplitude of the control test reflex (Crone, Hultborn, Mazieres, Norin, Nielsen, & Pierrot-Deseilligny, 1990). This aspect of H-reflex recruitment makes it difficult to determine the "true" amount of facilitation or inhibition. That is, a small control reflex could result in a large amount of facilitation, while a larger control reflex would result in a smaller amount of facilitation (Crone et al., 1990). Therefore, using percentage change, it may be difficult to estimate the real amount of either excitation or inhibition of the motor neuron pool.

The H/M ratio may be used as an alternate method of

analyzing changes in motor neuron reflex excitability The H/M ratio represents the percentage of the motor neuron pool that may be considered to be active at any particular time. The maximum N-response represents 100% of the activity of the specified muscle's motor neuron pool. The H/M ratio allows a quantification of the percentage of the active motor neuron pool. This interpretation enhances that ability to compare changes in the H-reflex between subjects (Crone et al., 1990). Crone and colleagues (1990) established that regardless of differences in the maximum H-reflex, the change in sensitivity to facilitation followed the same pattern as long as the same control reflex sizes were explored.






68



The level of maximum voluntary contraction has been shown to affect the H-reflex. Butler, Yue, and Darling (1993) evaluated isometric contractions up to 100% of a maximum voluntary contraction. The HIM amplitude of the soleus muscle was shown to be positively correlated to isometric contractions up to 50% of the maximum voluntary contraction. When the maximum voluntary contraction was 60% or above there was no correlation with force output and the HIM amplitude. In fact, the amplitude of the H-reflex showed no further increase. The H/M ratio was larger during rising rather than falling torque levels.

The relationship between force output and changes in the amplitude of H-reflex may be a result of different methods of motor unit recruitment at different force levels (Butler, Yue, & Darling, 1993). The soleus muscle, when producing a plantar flexion movement with the knee bend, should be considered the prime mover during the response. Therefore, the soleus muscle may be the prime mover up to 50% of a plantar flexion force (Hof & van den Berg, 1977). Up to 50% of a maximum voluntary contraction there was a linear increase in soleus muscle activation. When force output was greater than 50% of a maximum voluntary contraction, the gastrocnemius appeared to have an increased contribution on force output. The soleus motor neurons appeared to be mainly recruited up to 50 to 60% of a maximum voluntary contraction (Butler, Yue, & Darling, 1993).






69



Motor Neuron Excitability and Force


Changes in motor neuron pool excitability may relate to both force and timing of the movement (Frank, 1986). Butler, Yue, and Darling (1993) established that the amplitude of the H-reflex rose linearly with increased force output. This occurred up to 60% of a maximum voluntary contraction. After 60% of a maximum voluntary contraction, the H-reflex increased in curvilinear fashion. Corcos et al. (1990) indicated that the SS and ST strategies produce different excitation pulses, which should be reflected by changes in EMG activity and limb trajectory. Since the H-reflex rose in a curvilinear fashion after 60% of a maximum voluntary contraction, the relationships between EMG activity and limb trajectory may be compromised with large force productions. In fact, some of the tasks in the isometric study by Corcos et al. (1990) exceeded 60% of a maximum voluntary contraction. This may be a possible explanation for the lack of conclusive results. Further research is indicated to evaluate the relationship between the excitation pulse and ENG activity and peak torque in isometric contractions. The excitation pulse may additionally influence reaction times, which may be fragmented into premotor, motor and movement times.






70



Reaction Time Fracrmentation


The effects of task and stimulus complexity on movement parameters has been analyzed with premotor and movement times during simple reaction time paradigms. Premotor reaction time has been defined as the time from stimulus onset until onset of EMG activity. This period represents stimulus identification, response selection, and the transfer of the motor program to the appropriate muscle for movement execution (Schmidt, 1988). Motor time can be defined as the interval between the onset of EMG activity to initiation of the response which reflects the execution of the motor program (Eichenberger & Ruegg, 1984; Schmidt, 1988).

An increase in spinal excitability has been shown prior to the rapid generation of plantar flexion torque. This change has been used to further fragment premotor time into two separate intervals (Eichenberger & Ruegg, 1984). The

first interval (I,) was from the stimulus presentation until the onset of H-reflex facilitation. This interval may represent stimulus processing and response selection stages. The second interval (12) was from H-reflex facilitation until onset of EMG activity. This interval should reflect preparation of the motor system prior to movement execution. Any differences between these two intervals during different tasks should allow one to determine whether a change in programming or movement execution occurred (Eichenberger & Ruegg, 1984). These intervals were derived from changes in






71



spinal facilitation which occurred prior to EMG activity in voluntary movement.


Spinal Facilitation


Sullivan (1980) ascertained that the increase in spinal excitability occurred 40 ms prior to soleus EMG activity, while Mitchie et al. (1976) reported this increase some 90 ms to 120 ms prior to overt movement or 50 mns to 80 ms prior to EMG. According to Mitchie et al. (1976), the onset of spinal facilitation was time-locked to the onset of EMG and not to an initiation signal. That is, regardless of reaction time, the facilitation of the H-reflex always occurred within the 50 ms to 80 ms window prior to the onset of EMG activity. For reaction time tasks, more recent reports have confirmed these data (Brunt & Robichaud, 1996; Eichenberger & Ruegg, 1984; Ruegg, Krauer, & Drews, 1990; Riedo & Ruegg, 1988).

These findings appear to be in conflict with Frank

(1986), who reported that reflex facilitation was not timelocked to onset ENG but rather can be influenced by a subjects' preparatory set and duration of reaction time. The timing of reflex stimulation was shown to change in an anticipation timing task (Frank, 1986). In this case, the onset of facilitation was significantly earlier in relation to EMG activity onset for a coincident timing task than for a simple reaction time task. According to Gottlieb and colleagues (1989a) classification of movements based upon






72



the independent variables, this task would appear to be classified as an SI strategy. That is, a coincident timing task focuses on timing, not speed of the movement. Accordingly, this task would modulate the duration of the excitation pulse and agonist EMG activity bursts and inertial torques should rise at a uniform rate, independent of the movement speed. Frank's (1986) study seemed to support an increase in the pulse width (e.g. SI strategy). The timing of the reflex facilitation possibly reflected movement preparation. That is, movements that demand more precision (coincident timing versus simple reaction time) may require greater preparation time (increased pulse duration) between onset of reflex facilitation to premotor time.

The relationship between amplitude changes of the reflex facilitation may reflect an SS strategy which regulates pulse-height. This would account for the difference in Eichenberger and Ruegg's (1984) findings, in which shorter reaction times were associated with a greater amount of reflex facilitation. This would be attributed to these tasks being carried out using an SS strategy, while longer reaction times may have defaulted to the SI strategy. This could have been a complication of the actual design of the study. Even Eichenberger and Ruegg (1984) indicated that subjects may have used different movement strategies to perform the tasks.

The ability to differentiate the premotor period

according to movement preparation allows the analysis of the






73



motor set for a particular movement. The motor set implies the readiness of the central nervous system to implement a planned motor action (Brooks, 1986). For example, an increase in the period from reflex facilitation to onset of EMG activity could reflect a movement that requires precision of execution (Frank, 1986), whereas an increase in the premotor period prior to reflex facilitation could imply increased complexity in movement organization.



Motoneuron Reflex Excitability (Pulse-Width vs Pulse-Height)


Motor neuron pool excitability has been shown to

increase prior to the generation of plantar flexion torque (Brunt & Robichaud, 1996; Kaganffhara, KomLyaxna, Ohi, & Tanaka, 1990; Mitchie, Clarke, Sinden, & Glue, 1976; Sullivan, 1980). This increase occurred 50 mns to 80 mns prior to EMG activity onset in a non-choice ballistic task and has been shown to be time-locked to the onset of the ENG activity (Brunt & Robichaud, 1996; Kots, 1977; Mitchie et al., 1976). This period of spinal cord facilitation implies the readiness of the central nervous system to implement the planned motor action (Brooks, 1986; Frank, 1986). If the excitation pulse reflects spinal excitability, then changes in this period of spinal facilitation should correspond to changes in task or instructional variables.






74



There appears to be support for different task or

instructional variables being reflected by changes in spinal cord excitability. The dual-strategy hypothesis proposed that the SI strategy should be modulated by changing the timing (pulse width) of spinal cord facilitation. Kagamihara and colleagues (1990) evaluated the time course of spinal facilitation in a ramp plantar flexion task. In the ramp task, the subject moved at a preselected cursor speed. The onset of spinal facilitation occurred at 92 ms prior to EMG onset during the ramp task. Additionally, during an anticipation timing task there was a lengthening of the time period of spinal cord facilitation (70 ms prior to EMG onset) (Frank, 1986). The dual-strategy hypothesis proposed that the excitation pulse in an anticipation timing or ramp tasks should be classified as an SI strategy because timing and not speed was the dominate aspect of these tasks. Therefore, both of these examples appear to support pulse-width modulation of spinal cord facilitation.

Since speed was the dominate feature of the ballistic and step tasks, they would be classified as SS according to the dual-strategy hypothesis. Kagamihara and colleagues (1990) evaluated the time course of spinal facilitation in a step plantar flexion task. In this task, the subject was requested to reach a target that was designated between 20% to 40% of a maximum voluntary contraction. Spinal






75




facilitation occurred at 55 mns prior to EMG activity onset during the step task.

Numerous studies have established the onset on spinal facilitation to occur between 55 mns to 80 ms prior to EMG onset in ballistic plantar flexion tasks. However, shorter reaction times have greater amounts of spinal facilitation (Eichenberger & Ruegg, 1984). An analysis of this study showed a discrepancy in the time period of facilitation. That is, the period of facilitation for the short reaction times was 90 ins, while this time period was 180 mns for the longer reaction times. This appears to support the assumption that the short and long reaction times may have been carried out under different strategies, an observation which was supported by Eichenberger and Ruegg's (1984) suggestion that these differences may relate to different movement strategies. However, no study has been designed to analyze SS or SI strategies (e.g., pulse-width or pulseheight modulation) and modulation of spinal facilitation. In addition to different movement strategies, the amount of spinal facilitation may also be influenced by force production (Butler, Yue, & Ohi, 1993).


H-reflex Facilitation, Force and Reaction Times


According to Frank (1986), the degree of reflex facilitation may not only be related to timing of the movement but also to force and rate of force production.






76



Baba and Marteniuk (1983) identified a potential methodological problem in that allowing both peak force and time to peak force to randomly vary could well confound interpretation of the data. Previous studies indicated that the time needed to execute a movement can be influenced by both of these variables (Kawabe-Himeno, 1993; Ito, 1990, 1991; Meyer, Smith, & Wright, 1982).

The relationship between peak force and rate of rise of force has been described by the impulse variability model to the extent that when one variable was regulated, the other variable was proportionally rescaled (Brooks, 1986; Meyer et al., 1982). Although Eichenberger and Ruegg (1983) analyzed the relationship between the size of the H-reflex and movement time and amplitude, they reported inconsistent data for four subjects. These individual differences could well be due to different movement strategies. That is, did their subjects regulate peak torque or time to peak torque? Further research is indicated to determine the relationship between motor neuron excitability and different types of voluntary movements.



Summary


Single-joint isotonic movements can be classified

according to the dual-strategy hypothesis. There is some ambiguity as to the ability of this hypothesis to be extended






77



to isometric movements. The dual-strategy hypothesis proposes that movement control can be determined by the nature of the orientation of the instructions to perform a task. Different instructions may control movements by influencing the excitation pulse. This pulse, which may modulate spinal excitability, should be reflected by changes in the H-reflex. My study will examine isometric movements to determine if the dual-strategy can be expanded to account for these movements. The proposed neural mechanism (spinal excitability), that may govern the different strategies, will be evaluated.















CHAPTER 3
METHODS

Participants


Participants included 14 healthy individuals (8 males and 6 females) over the age of 18 years, who had no history of lower extremity disorders that would affect their participation in this study. However, only 11 participantss were included in the analysis. One participant was excluded because of inconsistent H-reflex responses, one due to nonequal pre and post maximum M-responses and one since the M-responses were inconsistent during testing. All participants had normal or corrected-to-normal vision and were asked to refrain from caffeine and alcohol 12 hours prior to the study (Eke-Okoro, 1982). Each participants dominate leg, which was designated as the preferred kicking leg, was tested.

Participants were recruited from colleagues, friends, and students from the Departments of Physical Therapy and Exercise and Sport Sciences. Self disclosure was utilized when determining if a participant met criteria for selection. All participants signed an informed consent form prior to testing (See Appendix A). IRB approval was obtained prior to testing (See Appendix B). All testing took place in the Physical Therapy motor Behavior lab.



78






79



Selection of Force Levels and Bandwidths



In a previous study, participants performed isometric plantar flexion contractions of 20, 40 and 60% of a MVC (Monohar, Brunt, & Robichaud, In Review). In this isometric condition, changes in MVC were equated to changes in distance utilized in the movement tasks. Isometric contractions were performed to bandwidths of 4, 8, 12, 16 and 20% of a participant's MVC. Results showed that participants utilized an SS strategy between the 8% and 12% bandwidths, while using an SI strategy within the larger bandwidths. Figure 3-1 illustrates a 40% MVC where the initial slope of force was similar for both the 4% and 8% and the 12%, 16% and 20% bandwidths. The switch between strategies was shown to occur between the 8% and 12% bandwidths. These results were also consistent for the 20% and 60% force level. Additionally, participants were shown to adopt an approximation of an SI strategy when they were instructed to use different percentages of a MVC to reach the same target bandwidth, which is shown in Figure 3-2. This figure demonstrates that the initial slope of force was similar when moving to the 8% bandwidth for the 20%, 40% and 60% force levels. These results were also consistent for the 12%, 16% and 20% bandwidths.






80










Force










L4,% L12 016% 20%









Figure3-.,. These traces represent an individual subject's response to a 40% force level to five different bandwidths.







81






Force











J//

t





-- 20%
40
-60%


Fiqrp 3-2 Teeta srpesn nidvda ujc' response to an8-adit ttre ifrn oc ees






82



The results of the above study were utilized to select the force levels and bandwidths for the current study. Twenty-five percent and 50% force levels were selected since no difference was noted in the initial EMG burst (Q30) when going to the same bandwidth, but moving between the 20%, 40% and 60% force levels. The 5% and 15% bandwidths were selected since a clear difference was observed in the initial slope of force between the 8% and 12% bandwidths.


Instrumentation


EMG

Electromyographic activity was recorded using surface electrodes on the dominate leg. The recording electrodes consisted of two silver-silver chloride electrodes 1 cm in diameter embedded in an epoxy-mounted preamplifier system (amplification x35) (Therapeutics Unlimited), the centers of which were 2 cm apart. A reference electrode was attached on the anterior surface of the tibial crest. The EMG signal was high (4,000 Hz) and low (20 Hz) pass filtered. The frequency response of the EMG signal was between 20 Hz to 4,000 Hz with the low-pass filter time constant being set at 2.5 ms. The signal was further amplified to give a final amplification between 5k to 10k. The processed signal was sampled on-line via an analog to digital converter (Biopac Systems Inc.) at 1000 samples per second for 2 minutes.






83



Surface electrodes were placed over the soleus,

gastrocnemius, and anterior tibialis muscles of the dominate leg. Soleus muscle activity was recorded over the dorsal medial surface of the posterior calf, inferior to the muscle belly of the gastrocnemius (Maryniak & Yaworski, 1987). Gastrocnemius activity was recorded over the muscle belly of the gastrocnemius on the posterior calf, while tibialis anterior activity was recorded over the muscle belly of the tibialis anterior on the anterior/lateral portion of the upper calf (Cram, 1991).



Stimulation

H-reflexes were elicited using an isolation unit

attached to a Grass 44 stimulator. A one ms rectangular pulse was used to elicit stimulation. A ground electrode was placed under the stimulated thigh. All data was sampled online and stored on an external drive for further analysis.



Force Transducer

A gensico cell load with a range from 0 to one hundred pounds of force was utilized. Force output was measured in voltage and then converted to both the percentage of the participant's maximum voluntary contraction and then Newtons for data analysis. The measure of Newtons was derived from converting volts into pounds and then multiplying this figure by the constant of one pound equaling 4.44 Newtons.






84



Participant Positioning



Participants were seated in a modified chair, as

depicted in Figure 3-3, which had attached leg and head rests. A device was attached to the chair that, regardless of leg length, maintained the participant's hip in 90* of

flexion, knee in 120* of flexion, and ankle positioned at neutral (Hugon, 1973). The participant was positioned in the chair with his/her back resting against the chair, arms folded across his/her lap and the head positioned on a head rest designed to eliminate head movement. The tested thigh was secured to the chair by straps that were placed over the medial and distal portions of the femur. The foot was secured in an ankle high wire-foam binding which was secured to a wooden board. Underneath the wooden board, at the level of the metatarsal heads, a force transducer was positioned to measure the force produced during voluntary isometric plantar flexion.



Testing sessions



Participants were asked to perform isometric plantar flexion movements at two specified force levels to two specific bandwidths. Two separate studies were conducted with participants being tested on two separate days. Each day they completed half of the protocol for each of the






85







Li ht
F
Cornputer




Stimulator
EMG
Leads



p Force
Transducer







Fignurp 3-3 Design of Experimental Set-up.






86



studies. That is, one day involved the analysis of isometric contractions performed to either 25% or 50% of a participant's maximum voluntary contraction. The second day involved the same analysis; however, the other force level was tested. These contractions were performed to bandwidths of 5% and 15% of the specified force level. The order of force levels (e.g., 50% and 25%) and bandwidths (e.g., 5% and 15% of the specified force level) was randomized to prevent any order effects.



Procedures



Study 1


Participants were instructed to perform 10 maximum

isometric plantar flexion contractions. The average of these contractions was used to calculate each participant's maximum voluntary plantar flexion contraction. Preselected percents (25% and 50%) of maximum voluntary contraction were calculated. Two bandwidths (5% and 15%) of the preselected percent maximum voluntary contraction were calculated. Reference lines corresponding to the specified bandwidth were shown on the computer screen. For example: The participant may have been asked to push down with an isometric contraction that was 25% of his or her maximum voluntary contraction within a 10% range. Therefore, reference lines






87



would be indicated at the levels of 20% and 30% of maximum voluntary contraction.

Testing began with learning trials to familiarize the participant with the testing procedure, task, and equipment. Participants practiced matching isometric plantar flexion contractions to the specific force level (25% or 50% of their maximum voluntary contraction). Concurrent feedback was used by having them view, in real-tine, their force output on a 35.56 cm computer screen placed 60 cm at eye level in front of them. output from the force transducer was sampled online on a computer screen and recorded for further analysis. After 50 trials the participant was asked to complete 10 additional trials. If 80% of the trials matched within the target bandwidth, the participant was considered to have learned the task and testing was initiated. During testing, participants received concurrent feedback. Additionally, these trials were used to calculate a participant's average premotor time for each specified force level.

A 5 mm red light emitting diode (intensity 30 to 40 mcd) was situated 60 cm in front of the participant. This was used as an initiation signal. Participants were instructed that following a red light signal to isometrically plantar flex their foot, which moved a cursor that measured force output on the computer. Participants were asked to perform isometric contractions to each specified force level (25% and






88



50% of the maximum voluntary contraction) to two target bandwidths (5% and 15% of the specified force level).

Participants did not exhibit any background EMG

activity. This was monitored on-line. Trials were repeated when a participant did not achieve the criterion force within the designated bandwidth. Participants were eliminated from the study when less than 80% of their trials did not met the specified criteria. No participant was eliminated from the study due to not meeting the above criteria.



Study 2


The sane protocol was followed in Study #2 as in Study #1, with the exception that H-reflexes were recorded during the testing session. The H-reflex was recorded from the surface electrode placed over the soleus muscle. One ms stimulations were applied to the skin over the tibial nerve in the popliteal fossa. Proper cathode placement was determined when 1) the direct motor reflex (M-wave) and Hoffman reflex (H-wave) displayed similar wave configurations, 2) the H-reflex was evoked before the Mreflex, and 3) the least amount of current was required to elicit a H-reflex. The stimulating intensity was increased until a maximum M-responses were observed, after which the average of 10 M-responses were calculated. The intensity was decreased until a 30% H-reflex, relative to the maximum M-






89



response was obtained. This was the stimulating intensity for the study. Ten baseline H-reflexes were recorded prior to testing.

Each participant completed six blocks of 15 trials which included tibial nerve stimulations that were 15, 30, 45, 60, 75 and 90 ms prior to the recorded average premotor time for the specified force level. The intervals within each trial block were randomized. A variable rest period of between 5 to 10 s was used between trials. After the trial blocks were completed, 10 maximum M-responses were again obtained.



Des icin and Analysis



The Acknowledge software program was utilized for data collection and analysis.


Study 1


The dual-strategy hypothesis was evaluated to determine if it could be extended to single-joint isometric movements. The dual-strategy hypothesis states that isotonic movements can be classified according to either a SS or SI strategy. The hypotheses tested was that single-joint isometric movements can be classified according to SS and SI strategies.

Dependent variables included the average of 7 trials for peak force, and the magnitude of the soleus EMG response






90



(integral of the rectified signal). Soleus EMG activity included the first 30 ms of activity and the activity from initiation until peak velocity. For data analysis, these values were normalized to the integral of a 100 ms maximum voluntary soleus contraction. All of these values were first divided out by there respective time. For example, the normalized value for Q30 was obtained by the formula (integral for Q30/30 ms) /( integral for 100 ms of MVC/ 100 ms) X 100. The final value obtained indicated the percentage of maximum of soleus EMG activity present. Additionally, the first time derivative of force which included 1) peak of the first time derivative (force-time velocity), 2) the time until peak velocity, and 3) the slope of the force-time velocity curve. Independent measures were bandwidths and force level. A 2 X 2 (Force X Bandwidth) repeated measures design ANOVA was used to compare differences among the test conditions on each dependent measure. All levels of significance was designated at R < 0.05. A Super ANOVA statistical program was used for data analysis.


Study 2


Different movement strategies were proposed to produce

different neural excitation pulses. This neural pulse may be reflected by changes in motor neuron reflex excitability. The hypotheses tested was that single-joint isometric movements performed under the SS and SI strategies produced






91



different intensities and patterns of motor neuron reflex excitability prior to EMG activity onset.

The dependent measures were the peak of spinal

facilitation at the onset of EMG activity, onset of change in spinal excitability, and the final slope of spinal facilitation. The onset of change in spinal excitability was determined by the time of the first data point which was one standard deviation above the baseline average. changes in the HIM ratio were used to calculate the dependent measures. The HIM ratio was calculated from the ratio of the respective (test) H-reflex to the average maximum M-response. The maximum M-response was the average of 10 responses. The HIM ratio was ranked in ascending order from the time of the elicited H-reflex to the onset of soleus muscle activity. once these ordered responses were rank ordered, a running average over 10 consecutive reflexes was calculated. That is, the average ordered trials 1 to 10 constituted the first value, while the second value was the average of trials 2 to 11. This running average was calculated for all the ordered H/M ratios for each force level and bandwidth. Using the running average, the dependent measures (slope, time and peak facilitation) were calculated for each force level and bandwidth.

Independent measures were specified force levels (25% and 50%) and bandwidths (5% and 15% of the specified force). A 2 X 2 ANOVA was used to determine any differences between