Preservation of force output through progressive reduction of stimulation frequency in human quadriceps femoris muscle.

Neuromuscular electrical stimulation (NMES) has long been used by physical therapists to maintain or increase the force-generating capacity of human skeletal muscle. Recently, numerous studies [1-13] have been undertaken to evaluate the effectiveness of NMES as a means of enhancing muscle force output. The majority of studies have used the peak torque generated during an isometric contraction to evaluate the effect on the muscle being studied and have shown that NMES can significantly increase the force-generating capacity of skeletal muscle in both healthy subject and patient populations. Although reported gainsappear to be positively correlated with the training intensity, [1,3,11] few studies have systematically varied the stimulation characteristics to identify which characteristics would allow maximum force generation during each training session. [14,15]

One important characteristic to consider when using NMES is the pulse frequency. The pulse frequency helps to determine the strength of contraction and the rate of force fatigue. The relationship between the pulse frequency and the force generated by a muscle has an approximately sigmoid shape. That is, low frequencies (eg, 1-10 pps) produce low force levels in the human quadriceps femoris musculature, whereas high frequencies (eg, 60-100 pps) are required to produce maximum force (SA Binder-Macleod, T Guerin; unpublished observatons; 1988). This relationship has been best described in cats, [16-19] but is also true for human skeletal muscle. [20] Force fatigue is defined as a decrease in the force-generating capacity of muscle. [21] The higher the stimulation frequency, the greater the rate of fatigue. [22,23] Therefore, a higher stimulation frequency (eg, 60 pps) may produce high force outputs, but will also produce a rapid rate of fatigue when compared with a lower stimulation frequency (eg, 20 pps). Conversely, a lower stimulation frequency will produce less fatigue than a higher frequency, but will also produce less force. [22] If increases in the force-generating capacity of a muscle are related to training intensity, stimulation patterns that maximize the force output by minimizing fatigue should be most effective. [24]

To evaluate the efficacy of NMES for increasing muscle force output, most clinical studies reported in the literature have stimulated the quadriceps femoris muscle, with frequencies ranging from 20 to 50 pps. All studies have used a single frequency throughout the entire training session; that is, no studies have changed the frequency during a treatment. In contrast, the discharge frequency of motor units has been shown to decline during fatiguing contractions in human subjects. [25]

Human skeletal muscle, when stimulated at the frequency needed to produce maximum force, displays a much higher fatigue rate during continuous 60- or 90-second contractions than during similar-duration voluntary contractions. [22,23] Low frequency stimulation (eg, 20 pps) results in little force fatigue, but the initial production of force is only about 65% of the force generated by high frequency (eg, 50-80 pps) stimulation in the unfatigued state. [22] The rapid loss of force during NMES can be significantly reduced by a progressive reduction in stimulation frequency over time. [22,23] Jones et al [22] markedly reduced the rate of force fatigue in the human adductor pollicis muscle by progressively reducing the stimulation frequency from 60 to 20 pps during a continuous, 60-second contraction. The fatigue rate observed with progressive reduction in stimulation frequency closely matched that of a maximal voluntary contraction. Marsden et al [23] referred to this procedure as "artificial wisdom." [23]

Clinically, in contrast to the continuous stimulation used by Jones et al, [22] intermittent stimulation of skeletal muscle is used to strengthen muscle. Duty cycles ranging from 10 to 15 seconds "on" and 8 to 50 seconds "off" have generally been used. The purpose of this study was to compare the rate of fatigue during intermittent NMES using a protocol with a progressive reduction of stimulation frequencies with the rate of fatigue using a protocol with a consistent stimulation frequency. We hypothesized that a progressive reduction in stimulation frequency would be capable of reducing the rate of fatigue during intermittent stimulation that used a train duration similar to those used clinically to strengthen muscle.

Method

Subjects

Twelve healthy subjects (6 men, 6 women), ranging in age from 19 to 37 years (X = 22.1), volunteered to participate in the study. All subjects signed informed consent statements and completed the study. The subjects reported no history of lower extremity injury. Subjects were not permitted to participate in any strenuous exercise for 24 hours prior to each experimental session.

Procedure

Each subject's right quadriceps femoris muscle was electrically stimulated with a series of 350-[micro]sec, squarewave, monophasic pulses to produce isometric contractions. This pulse duration was long enough to provide sufficient pulse charge to produce strong recruitment, yet short enough to allow the stimulation to be generally well-tolerated by the subjects. A constant-voltage stimulator, (*) controlled by an IBM-PC computer, (+) was used to generate each pulse. The computer was equipped with a Data Translation 2806 Counter-Timer Board, (**) and custom-made software allowed the exact timing of all aspects of each experimental protocol. The amplitude of the pulse was set to obtain a strong contraction by increasing the voltage to subject tolerance or until 100 V (maximum output of the stimulator) was reached. Subjects tolerated voltages ranging from 70 to 100 V. This intensity produced sub-maximal contractions in all subjects. Pulses were grouped into trains that lasted 8 seconds. Each train was followed by a rest period of 12 seconds. A short rest time was purposely used to produce the greatest amount of fatigue in the shortest amount of treatment time. [15] This pattern of 8 seconds on/12 seconds off was repeated 30 times for a total of 10 minutes. One experimental protocol (constant-frequency protocol) used the same stimulation frequency for each train. The other experimental protocol (variable-frequency protocol) progressively reduced the stimulation frequency across trains. That is, we did not change the frequency within any train, but we did decrease the stimulation frequency of successive trains.

The minimum stimulation frequency that produced maximum force in an unfatigued muscle (ie, 60 pps) was used in the constant-frequency protocol, and the frequencies that best preserved force and minimized fatigue were used in the variable-frequency protocol. The specific frequencies used were determined during pilot work, which was conducted with two subjects who were also participants in this study. All contractions throughout the study were isometric.

Custom-made electrodes were used to deliver the stimulation to the targeted muscles. The electrodes were rectangular (15.3 X 15.7 cm) with rounded edges, 0.1 cm thick, and made of aluminum. Stimulation using these electrodes was well-tolerated by our subjects and allowed strong levels of contraction. One electrode (anode) was placed over the proximal anterior thigh just distal to the inguinal ligament, and the other electrode (cathode) was placed with its distal edge 5 to 7 cm proximal to the superior border of the patella. The electrodes were bent to conform to the contour of the subject's thigh. A thin sponge, thoroughly soaked with water, was placed between the electrode and subject to serve as the conducting medium. The electrodes were secured to the subject's thigh with elastic straps.

The force output of the contractions produced by electrical stimulation were recorded using the KIN-COM[TM] II dynamometer. [Section] The KIN-COM[TM] II is a computer-controlled dynamometer that performed analog-to-digital conversion and storage of the force data with a sampling frequency of 100 samples per second. All data were stored on the hard disk of the personal computer used to control the dynamometer. The subjects were seated on the dynamometer with their right knees positioned at a joint angle that produced no passive force. Knee joint angles ranged from 80 to 85 degrees of flexion. Subjects' hips were positioned in approximately 70 degrees of flexion. Stabilization was provided by a back support, lap seat belt, and thigh strap. The rotational axis of the dynamometer was aligned with the knee joint axis (ie, lateral epicondyle of the femur). The lower edge of the force transducer pad was positioned against the anterior aspect of the subject's leg, about 5 cm proximal to the lateral malleolus. The exact position of the dynamometer head, seat, thigh support, and force transducer was recorded for each subject to ensure consistency across sessions.

Pilot Work

To identify the frequency to be used during the constant-frequency protocol, we first plotted the force-frequency relationship of the quadriceps femoris muscle. Sixty pulses per second was determined to be the minimum frequency that would produce maximum force and therefore was used for the constant-frequency protocol. The frequency characteristics of the variable-frequency protocol were next determined. Each subject was stimulated with a series of trains with a duty cycle identical to that of the constant-frequency protocol. That is, trains of 8 seconds on/12 seconds off were repeated a total of 30 times. As with the constant-frequency protocol, the frequency of the first train was 60 pps. The stimulation frequency was then reduced between successive trains to determine which pattern of stimulation best preserved force and minimized fatigue. Approximately five different stimulation frequencies were tried for each subject. Each subject was tested with only one stimulation frequency per test session, conducted on separate days. The method for selecting the frequencies to test and eventually use in the study was primarily empirical. The selected protocol lowered the stimulation frequency in 5-pps steps for trains 2, 3, 5, 8, 12, and 20. Trains 20 to 30, therefore, were delivered at a frequency of 30 pps.

Data Collection

After determining the appropriate frequencies to use for each experimental protocol, the actual data-collection procedure was begun. Approximately 2 months elapsed between the conclusion of pilot work and the commencement of data collection. Each of the 12 subjects (including the 2 pilot subjects) participated in three data-collection sessions. During the first session, the subjects were familiarized with the experimental protocol and the stimulation voltage to be used for the following experimental sessions was determined. The same voltage was used for both experimental sessions.

After the familiarization session, each subject participated in one constant-frequency session and in one variable-frequency session. The order of these two sessions was randomly determined for each subject. The variable-frequency session was identical to the constant-frequency session, except that the stimulation frequency was progressively reduced between trains in the variable-frequency session. During both sessions, several brief (1-second) contractions were elicited before testing to ensure proper functioning of equipment. No fatigue was observed as a result of these contractions. At least 24 hours separated each session, but most subjects waited 3 days between sessions because of the fatigue and soreness experienced after the first session. Subjects were instructed to refrain from any voluntary effort during any of the electrically elicited isometric contractions.

Data Analysis

The average force over each 8-second contraction was calculated using the KIN-COM [TM] II's evaluation program, and all statistical tests were performed using a personal computer and a Lotus 1-2-3 spreadsheet. (1) All analyses were performed on normalized force data. To normalize the force recordings, the average force produced by each contraction was divided by the average force of the initial 8-second contraction elicited during that session (ie, 60 pps for both sessions). Normalization of forces helped to reduce the variability across subjects and within subjects across sessions. The differences between the mean normalized force values produced by the constant-frequency and variable-frequency protocols were compared for statistical significance using a one-way, repeated-measures analysis of variance (ANOVA), followed by two-tailed Student's t tests repeated for each 8-second contraction.

Reliability was determined by comparing the average force outputs of the first contraction for each experimental session. For both sessions, identical stimulation frequencies were used (60 pps). Intraclass correlation coefficients (ICCs) were determined using a repeated-measures ANOVA and found to be .778 for all treatments (ICC[1,k]) and .636 for a single treatment (ICC[1,1]) (F = 4.50, P < .007). Given the relatively small sample size (N = 12), we concluded that our measurements were sufficiently reliable.

Results

The Figure shows that the decline in force of both stimulation protocols was approximately equal over the first 3 contractions (data plotted at 0, 20, and 40 seconds); for the next 4 contractions, the rate of force fatigue decreased for the variable-frequency stimulation. No similar decline in the rate of fatigue was observed when the constant-frequency protocol was used. For the remaining 23 contractions, the rate of force fatigue was approximately equal for both stimulation protocols. Between the 8th and 30th contractions, the mean difference in average normalized force output between the two groups remained approximately 0.16. Because of the continued fatigue produced by both protocols, however, the percentage of difference in force produced by the variable-frequency protocol compared with the constant-frequency protocol (%DIFF) increased from 31% for the 8th contraction (140 seconds) to 46% for the 30th contraction (580 seconds). The %DIFF was calculated by taking the difference in the mean force produced by the variable-frequency protocol and the constant-frequency protocol, dividing this difference by the mean force of the constant-frequency protocol, and then multiplying this quotient by 100.

Table 1 compares the results from the constant-frequency and variable-frequency protocols. The one-way, repeated-measures ANOVA revealed a significant difference in force outputs between the two protocols (F = 4.89, P < .01). The post hoc paired t tests revealed that by the sixth contraction the difference in average force between the two groups was significant (P < .05) and, except for the ninth contraction, remained significant for the remaining 23 contractions.

Table 2 compares the results for each subject's 30th contraction (580 seconds) for each protocol. The 30th contractions were compared because, as shown in Table 1, this contraction produced the greatest fatigue using either protocol. The %DIFF varied considerably, with a range of -6% to 132%. The variable-frequency pattern was tested and developed on subjects

[TABULAR DATA OMITTED]

1 and 2, who demonstrated %DIFFs of 60% and 81%, respectively. The %DIFFs of subjects 1 and 2 were well above the mean %DIFF of all 12 subjects for the 30th contraction (X = 46%).

Discussion

The purpose of this study was to determine whether decreasing the stimulation frequency during intermittent fatiguing contractions would decrease the fatigue rate, as compared with using a constant stimulation frequency. As anticipated, a progressive reduction in stimulation frequency across trains resulted in a decrease in force fatigue. This finding supports those of Jones et al [22] and extends their findings to include intermittent contractions.

Fatigue is the result of failure, at one or more sites, of the chain of events

[TABULAR DATA OMITTED]

leading to muscle contraction. [26] The exact location of the failure is a function of the stimulation characteristics and the physiological properties of the muscle being stimulated. Marsden and colleagues [23] have shown that the amount of muscle fatigue is directly related to the number of pulses received by the muscle. Therefore, in general, the greater the pulse frequency, the more rapidly fatigue develops.

Fatigue is generally grouped into two categories: "electrical fatigue" and "mechanical, or metabolic, fatigue." [26] Failure to produce a muscle contraction prior to the release of calcium from the sarcoplasmic reticulum is generally considered to be electrical fatigue; subsequent failure to produce a muscle contraction is generally considered to be mechanical, or metabolic, fatigue. Electrical changes that affect the amplitude of the action potential along the sarcolemma will also affect the subject's electromyographic (EMG) activity. Therefore, when EMG activity and force decline concurrently, the fatigue is generally described as being electrical in origin. If the decline in force precedes the decline in EMG activity, then the fatigue is thought to be caused by mechanical, or metabolic, deficiencies. Problems in excitation-contraction coupling, although electrical in nature, may not affect the EMG signal. Both processes, electrical and mechanical fatigue, can occur simultaneously. [27]

Bigland-Ritchie et al [27] suggested the force fatigue produced by high frequency stimulation results in part from failure in electrical propagation along the muscle fiber membrane. Electrical fatigue is generally not seen during voluntary contractions. [27] The rapid loss of force produced by the electrical stimulation of muscle is largely reduced during voluntary contractions by a progressive reduction in motor-unit firing frequencies. [22,23,25] Therefore, the decreased fatigue rate observed during stimulation with the decreasing-frequency protocol may be due to the decreased electrical fatigue produced by the lower stimulation frequencies. No EMG recordings were made during this study; therefore, any explanation of the causes of fatigue or mechanisms for reducing fatigue by decreasing the stimulation frequency is only speculative.

Relatively few studies investigating muscle strengthening in human subjects with electric stimulation have examined the fatigue rates during training. Currier and Mann [3] reported a decrease in force of 21% after 10 contractions elicited by constant-frequency intermittent stimulation in contrast to 50% force loss after the 10th contraction (constant-frequency protocol) in our study. This difference may be due to differences in the stimulation characteristics used. Currier and Mann [3] used an Electrostim 180-2, (#) which uses sine waves at a frequency of 2,500 Hz, delivered with 10-msec "on" bursts at a frequency of 50 bursts per second. Although the duty cycle in the study of Currier and Mann [3] included a longer period of stimulation (15 seconds) compared with our study, the rest time in their study also was significantly longer (50 seconds). This longer period of rest may have allowed for greater force recovery. [15]

Selkowitz, [11] using an identical stimulator to that used by Currier and Mann, [3] found no fatigue during training of the subjects in his study. A 2-minute rest period was allowed between each contraction to help minimize fatigue. Two additional stimulation characteristics, which were varied during the study of Selkowitz, [11] make comparison of fatigue rates between our studies difficult. First, the contraction time ("on" time) in Selkowitz's [11] study was limited to the onset of fatigue. Therefore, if the muscle fatigued more rapidly during the course of treatment, the contraction duration was shortened; this limited contraction time would have minimized any changes in the mean torque production. Second, and more significant, the stimulation intensity in that study was not kept constant during training, but instead was increased throughout the treatment to maintain torque output. The increase in intensity would have recruited previously inactive motor units. This is a very different situation than that created in our study, where intensity was kept constant to attempt to recruit a constant motor-unit population. Therefore, factors other than just the rest time between contractions may account for the very diffrent fatigue rates observed.

Because the same protocol was used for all subjects in our study, the optimal stimulation protocol may not have been used for each subject. This point is supported by the range of responses observed. The two subjects who were used to develop the variable-frequency protocol showed greater %DIFFs than the group mean. Several subjects responded even more favorably (ie, subjects 5, 8, and 11), whereas others showed much smaller %DIFFs (ie, subjects 3, 6, and 9) (Tab. 2). We also must acknowledge the fact that the constant-frequency protocol may not have been optimal for all subjects. This fact may have contributed to the wide range of forces observed at the end of the constant-frequency protocol (19%-48% of initial force), as well as the range of differences between the two protocols. Further studies are needed to identify the variable-frequency protocol that maximizes force production for each subject. Further studies are also needed to investigate whether decreasing the force fatigue during a training session through the use of a variable-frequency stimulation protocol can actually increase the gains in force output compared with constant-frequency stimulation.

Conclusion

The results of this study indicated that, compared with constant-frequency NMES, variable-frequency stimulation can markedly decrease the rate of force fatigue in skeletal muscle during intermittent contractions. Further investigation is needed to optimize the stimulation characteristics and to determine whether variable-frequency stimulation will also result in increased muscle strengthening. The results of this study should encourage clinicians to attempt to identify the best stimulation frequency to use with each of their clients.

(1) Lotus Development Corp, 55 Cambridge Pkwy, Cambridge, MA 02147.

(#) Micromed Instruments, 4996 Place de la Savane, Montreal, Quebec, Canada H4B 1R6.

References

[1] Boutelle D, Smith B, Malone T. A strength study utilizing the Electro-stim 180. Journal of Orthopaedic and Sports Physical Therapy. 1985;7:50-53.

[2] Currier DP, Lehman J, Lightfoot P. Electrical stimulation in exercise of the quadriceps femoris muscle. Phys Ther. 1979;59:1508-1512.

[3] Currier DP, Mann R. Muscular strength development by electrical stimulation in healthy individuals. Phys Ther. 1983;63:915-921.

[4] Delitto A, Rose SJ, McKowen JM, et al. Electrical stimulation versus voluntary exercise in strengthening thigh musculature after anterior cruciate ligament surgery. Phys Ther. 1988;68:660-663.

[5] Duchateau J, Hainaut K. Training effects of sub-maximal electrostimulation in a human muscle. Med Sci Sports Exerc. 1988;20:99-104.

[6] Kramer JF, Mendyrk SW. Electrical stimulation as a strength improvement technique: a review. Journal of Orthopaedic and Sports Physical Therapy. 1982;4:91-98.

[7] Mohr T, Carlson B, Sulentic C, Landry R. Comparison of isometric exercise and high volt galvanic stimulation on quadriceps femoris muscle strength. Phys Ther. 1985;65:606-612.

[8] Laughman RK, Youdas JW, Garrett TR, Chao EYS. Strength changes in the normal quadriceps femoris muscle as a result of electrical stimulation. Phys Ther. 1983;63:494-499.

[9] McMiken DF, Todd-Smith M, Thompson C. Strengthening of human quadriceps muscles by cutaneous electrical stimulation. Scand J Rehabil Med. 1983;15:25-28.

[10] Romero JA, Sanford TL, Schroeder RV, Fahey TD. The effects of electrical stimulation of normal quadriceps on strength and girth. Med Sci Sports Exerc. 1982;14:194-197.

[11] Selkowitz DM. Improvement in isometric strength of the quadriceps femoris muscle after training with electrical stimulation. Phys Ther. 1985;65:186-196.

[12] Soo C-L, Currier DP, Threlkeld AJ. Augmenting voluntary torque of healthy muscle by optimization of electrical stimulation. Phys Ther. 1988;68:333-337.

[13] Stefanovska A, Vodovnik L. Changes in muscle force following electrical stimulation: dependence on stimulation waveform and frequency. Scand J Rehabil Med. 1985;17:141-146.

[14] Snyder-Mackler L, Garrett M, Roberts M. A comparison of torque-generating capabilities of three different electrical stimulation currents. Journal of Orthopaedic and Sports Physical Therapy. 1989;11:297-301.

[15] Packman-Braun R. Relationship between functional electrical stimulation duty cycle and fatigue in wrist extensor muscles of patients with hemiparesis. Phys Ther. 1988;68:51-56.

[16] Botterman BR, Iwamoto GA, Gonyea WJ. Gradation of isometric tension by different activation rates in motor units of cat flexor carpi radialis muscle. J Neurophysiol. 1986;56:494-506.

[17] Cooper S, Eccles JC. The isometric response of mammalian muscles. J Physiol Lond. 1930;69:377-385.

[18] Mannard A, Stein RB. Determination of the frequency response of isometric soleus muscle in the cat using random nerve stimulation. J Physiol Lond. 1973;229:275-296.

[19] Kernell D, Eerbeek O, Verhey BA. Relation between isometric force and stimulation rate in cat's hindlimb motor units of different twitch contraction times. Exp Brain Res. 1983;50:220-227.

[20] Binder-Macleod SA. Force-frequency relationship in skeletal muscle. In: Currier DP, Nelson RM, eds. Excitable and Connective Tissue: Recent Advances and Clinical Concepts. Philadelphia, Pa: FA Davis Co. In press.

[21] Edwards RHT. Human muscle function and fatigue. Ciba Found Symp. 1981;82:1-18.

[22] Jones DA, Bigland-Ritchie B, Edwards RHT. Excitation frequency and muscle fatigue: mechanical responses during voluntary and stimulated contractions. Exp Neurol. 1979;64:401-413.

[23] Marsden CC, Meadows JC, Merton PA. "Muscular wisdom" that minimizes fatigue during prolonged effort in man: peak rates of motoneuron discharge and slowing of discharge during fatigue. In: Desmedt JE, ed. Motor Control Mechanisms in Health and Disease. New York, NY: Raven Press; 1983:169-211.

[24] Delitto A, Snyder-Mackler L. Two theories of muscle strength augmentation using percutaneous electrical stimulation. Phys Ther. 1990;70:158-164.

[25] Bigland-Ritchie B, Johansson R, Lippold OCJ, et al: Changes in motoneuron firing rates during sustained maximal voluntary contractions. J Physiol Lond. 1983;340:335-346.

[26] Clamann HP. Fatigue mechanisms and contractile changes in motor units of the cat hindlimb. Can J Sport Sci. 1987;12:20s-25s.

[27] Bigland-Ritchie DA, Jones DA, Woods JJ. Excitation frequency and muscle fatigue: electrical responses during human voluntary and stimulated contractions. Exp Neurol. 1979;64:414-427.

S Binder--Macleod, PhD, PT, is Assistant Professor, Program in Physical Therapy, School of Life and Health Sciences, University of Delaware, 054 McKinly Laboratory, Newark, DE 19716 (USA). Address all correspondence to Dr Binder-Macleod.

T Guerin, BS, PT, is a physical therapist in the Johnson Rehabilitation Institute of the John F Kennedy Medical Center, James St, Edison, NJ 08817. He was a student in the Program in Physical Therapy, School of Life and Health Sciences, University of Delaware, when this study was performed in partial fulfillment of the requirements for his degree of Bachelor of Science in Physical Therapy with Distinction.

This work was supported by the University of Delaware Honors Program, by a Peter White Fellowship to Mr Guerin, and by grants from the University of Delaware Research Foundation and The Foundation for Physical Therapy Inc to Binder-Macleod.

This study was approved by the University of Delaware's Human Subjects Committee.