Effect of variation in the burst and carrier frequency modes of neuromuscular electrical stimulation on pain perception of healthy subjects.

Key Words: Electrotherapy, electrical stimulation; Muscle performance, lower extremity; Pain.

Use of neuromuscular electrical stimulation (NMES) to develop muscle strength (torque, force), to improve athletic performance, and to increase muscle size has led to intensified research interest in this modality as a rehabilitative and muscle performance enhancing tool.[1-11] Claims attributed to Kots of 100% increases in torque of untrained healthy subjects and 40% increases in torque of trained athletes served as the impetus for much of these investigations.[3]

Although training effects (increasing torque) have been achieved using NMES with training intensities ranging from 25%[12] to 91%[8] of Maximal voluntary contraction (MVC), many researchers[5,6,8,12] have proposed that volitional exercise intensities between 50% and 100% of the MVC are most effective for increasing torque.[13] For example, Lai and coworkers[12] showed that subjects training with NMES at 50% of MVC had higher gains of knee extensor torque than did those using NMES of only 25% of MVC. Muscle contractions induced by NMES at these intensities, however, are hindered by subject perception of pain associated with the stimulus,[14] and this is the major limitation to its clinical use for strength training.[2]

Investigators have found that they could vary subjects' perception of pain by varying the current frequencies or waveforms of the carrier wave used for electrical stimulation. Vodovnik and coworkers[15] found less pain with the use of sine-wave frequencies greater than 500 Hz, whereas Crochetiere and associates,[16] using a square wave, found frequencies of 300 pulses per second (pps) to be most comfortable. Wong[17] reported that twin-peaked monophasic pulses were less painful than asymmetrical biphasic pulses. Delitto and Rose[18] could not find subject preference for three different waveforms (sine, sawtooth, and square) using current characteristics of 2,500 Hz, 50 pps, and 10 milliseconds' pulse duration. Baker and coworkers[19] found subject preference for symmetrical biphasic square waves (32 pps) over sine waves at a 2,500-Hz frequency enveloped into 10-millisecond packages presented 50 times per second.

Kots[1] stated that in order for NMES to be effective, the stimulus must produce a relatively low pain level and the current must be of adequate magnitude and frequency to produce high intensity tetanic muscle contractions. Moreno-Aranda and Seireg[20] used an NMES unit that meets these pain and contraction requirements by delivering sine-wave stimuli at frequencies of 2,000 to 10,000 Hz. In order to produce maximum muscle contractions, they modulated (interrupted) the carrier frequency between 10 and 500 times per second.[20] This base pulsed rate was called a "carrier frequency" in that article (frequency of pulses within a burst[21]), and the modulated carrier frequency was called a "burst." Bursts are finite series of pulses having fixed amplitude, duration, and rate (Fig. 1).[22] Moreno-Aranda and Seireg[20] reported that modulated carrier frequencies of 2,500 to 5,000 Hz produced considerable torque and that frequencies of 9,000 and 10,000 Hz were associated with the least subject pain.

Kots[1,23] reported muscle contraction forces greater than voluntary contractions by using NMES with a modulated current that produced relatively little pain. Kots may also have considered the effect of tissue impedance on subject comfort when using 1,000 Hz[23] and then changing to a 2,500-Hz carrier frequency.[1] By increasing the frequency (reducing phase duration) of the NMES, the tissue opposition to current flow decreases to improve Comfort.[14,24] Kots probably believed that the problem of high intensity muscle contractions accompanied with pain was resolved by using a combination of burst frequencies (eg, 50, 70, and 90 bursts per second [bps]) and carrier frequencies (eg, 2,500, 5,000, and 10,000 Hz). The number of bursts per second, however, is physiologically the same as the number of monophasic or biphasic pulses per second in producing muscle contractions.[25]

The form of current used by Kots has been referred to as "Russian" current or technique and consists of a continuous sine-wave output of 2,500 Hz (serving as a carrier frequency) modulated to provide bursts, each of 10 milliseconds' duration, separated by interburst intervals of 10 milliseconds.[14] Several NMES units are marketed that produce "Russian" current, and some units permit physical therapist selection of burst and carrier frequency combinations for clinical use. Studies comparing combinations of burst and carrier frequencies are lacking in the literature. Information enabling proper selection of burst and carrier frequencies for effective patient management is needed. The purpose of this study was to ascertain the effect that various combinations of burst and carrier frequencies have on subjects' perception of pain intensity, while NMES produced a constant level of muscle torque. We expected to find a specific frequency combination (nine combinations of three burst frequencies and three carrier frequencies) to be best.

Method

Subjects

Twenty-seven healthy volunteers (22 men, 5 women), ranging between 18 and 35 years of age, participated in the study. The volunteers responded to announcements requesting subjects for the study. Subjects had no history of injury or known pathology of their knees and no previous experience with NMES. All subjects consented to participating in this study. Descriptive data of the subjects are shown in Table 1. All subjects in the study were able to accept sufficient NMES to cause 50% of their MVC and volitionally produce MVC within 75% of their body weight. One researcher group[26] has reported that MVC force of the knee extensors is 75% proportional to a subject's body weight (r=.92). Because the authors at the initial session found torque values of female volunteers to be similar to those of male volunteers, the female volunteers were included in the sample studied.

Table 1. Descriptive Data of Subjects (N=27)
             Weight     Height          Age         Torque(a)
Statistic     (kg)       (cm)           (Y)            (N-m)
X [bar]       71.5       176.0          22.0           269.6
SD             9.4         7.3           2.2            35.9
Range        54.5-93.1   160.0-193.4   18.0-35.0    146.4-244.0
(a) Knee extension torque  maximal voluntary contraction).

Procedure

Initial session. In the initial session, the subjects' height, weight, and age were recorded. Then subjects were familiarized with the test equipment and procedure. After familiarization, 5 minutes of rest to avoid fatigue was given each subject before engaging in the test portion of the session. Subjects' maximum isometric extensor torque of their nondominant knee (chosen arbitrarily) was determined using a Cybex[R] II dynamometer system.(*) Torque calibrations with the chart recorder set on damp 3 were performed on a regular basis.

Each subject was seated on the dynamometer chair with the backrest inclined to 60 degrees and the tested knee fixed at 60 degrees of flexion. This position has been shown to be useful for yielding maximum torques.[27,28] Body stabilization was provided during the torque determinations by a lap seat belt and by thigh and ankle straps, while subjects grasped the handles along the sides of the test chair. The axis of rotation of the dynamometer was aligned with the tested knee, and the dynamometer speed selector was set at 0 [degrees]/s. Subjects were asked to perform three consecutive isometric MVCs, each lasting about 3 seconds. The resultant torques were transcribed by a chart recorder. The highest peak torque reading of the three voluntary contractions was used as the MVC and for further analysis. Next, each subject was given sufficient time to become familiar with NMES while remaining seated on the dynamometer chair.

An Electrostim 180-2i unit([dagger]) was selected for the "Russian" current because it is commonly used clinically, can be used to produce ample torque,[29] and was developed to improve subject comfort.[9,19] This NMES unit delivers individual sine waves at variable carrier frequencies (2,500, 5,000, and 10,000 Hz) and burst frequencies (50, 70, and 90 bps). An example of current output characteristics is a 2,500-Hz carrier frequency (200-microsecond phase duration) modulated to deliver 50 bps, with each burst lasting 10 milliseconds and with bursts separated by interburst intervals of 10 milliseconds.[14] One of two surface stimulating electrodes (12x8-cm carbonized rubber) was placed longitudinally on the subject's skin over the femoral triangle, and a similar electrode was placed over the vastus medialis muscle.[30] Sponges moistened with tap water and placed between the skin and the electrodes (commonly used clinical method) served as a mean-impedance[31] conducting medium. Prior to stimulation, the electrode/skin impedance was recorded([double dagger]) to test the quality of the electrode attachment and skin preparation. An impedance level of less than 50,000 [Omega] was considered arbitrarily acceptable to proceed with the NMES testing. Electrodes were strapped securely to the thigh by elastic wraps.

Subjects were instructed to control the stimulus and time for their left knee extensors to produce a torque equivalent to 50% of their predetermined MVC, and to avoid volitional contraction of their quadriceps femoris and hamstring muscles during NMES. The tester (JGR) constantly observed these muscles and the shape of the torque curves for any tendency to volitionally contract them and to ensure 50% of MVC by NMES on all contractions.

Experimental sessions. Following the initial test session, three experimental sessions (sessions 2, 3, and 4) were held, with a 1-week interval between sessions. Each experimental session consisted of three NMES treatments. A treatment consisted of 10 muscle contractions at 50% of MVC, induced by NMES that delivered a randomly selected burst and carrier frequency combination (eg, 50 bps/5,000 Hz). Each induced contraction lasted 15 seconds (5-second ramp on) followed by a 50-second rest (23% duty cycle). Stimulus amplitude was adjusted by the tester about every two contractions to maintain torque at 50% of MVC (amplitude increased 6-20 mA over initial level). A treatment set of 10 induced contractions was followed by a 5-minute rest period, a period we believed to be sufficient to recover from any muscle fatigue.[32] This treatment sequence was then repeated two more times per experimental session with different randomly selected combinations of burst and carrier frequencies. An experimental session of three NMES treatments lasted 30 to 45 minutes (7.5 minutes of NMES), and the three experimental sessions (2, 3, and 4) involved a total of nine burst and carrier frequency combinations.

In the 5-minute rest period following the 10th induced contraction of each set of frequency combinations, the subjects rated their perceived pain intensity using a visual analog scale (VAS), a 10-cm horizontal line. The 10-cm line was labeled "no pain" at 0 (left end) and "pain as bad as it could be" at 10 cm (right end). The VAS provides a ratio-level measurement[33] that is reproducible[34-36] and correlates with other methods of measuring pain.[37] Twelve subjects were randomly selected in the initial session to rate their pain with a VAS after random measurements administered to assess the reliability of the instrument when measuring pain associated with NMES. A random model for intraclass correlation provided a coefficient of .954.[38] The time selected for administering the VAS negated recall or bias toward the current frequency combinations used in the study.

Data Analysis

The VAS was arbitrarily divided into low and high ratings by dividing the 10-cm line into halves. The frequencies of subject ratings occurring in the left half were counted to indicate the number of times a burst and carrier frequency combination was rated as most comfortable.

Descriptive statistics (mean, SD) were calculated for demographic and recorded scores. Analyses of variance for repeated measures (burst/carrier frequency) were used to analyze the VAS and peak current (milliamperage) data. An alpha level of .05 was used in the analysis. The Duncan post hoc analysis was used to test all pairwise means when significant F ratios occurred. Relations between milliamperage (read from the NMES unit meter) and pain intensity ratings were determined by the Pearson Product-Moment Correlation method.

Results

Table 2 discloses that combinations of burst frequencies (50, 70, and 90 bps) and carrier frequencies (2,500, 5,000, and 10,000 Hz) significantly influenced the pain intensity ratings of subjects receiving NMES in this study. All burst combinations with carrier frequencies of 2,500 and 5,000 Hz led to significant differences in pain from the burst combinations with the carrier frequency of 10,000 Hz. No statistical significance was found between the pain for combinations of burst frequencies (50, 70, and 90 bps) and carrier frequencies of 2,500 and 5,000 Hz. Figure 2 illustrates the trend of decreasing pain intensity ratings as the burst rate increased and of increasing pain intensity ratings as the carrier frequency increased in rate.

Table 2. Analysis of Variance Summary for Pain Intensity Ratings of Nine
Combinations of Burst and Carrier Frequencies
Source            df      SS       F         P
Burst/carrier
frequency          8    233.33    29.17     .005
Subjects          26    527.13    20.27
Interaction      208    632.10     3.04
Total            242   1392.56

Discussion

In addition to information revealed by the VAS data, the symmetry and consistency of torque curve patterns was disrupted by the burst combinations with the carrier frequency of 10,000 Hz. Figure 3 shows that the 10,000-Hz carrier frequency imposed the appearance of unfused tetanic muscle contractions on top of the recorded torque curve.

There was a marked variation in individual pain intensity ratings ranging from 0.2 to 9.6 cm on the VAS. The frequency of subjects selecting a burst and carrier frequency combination on the VAS on the left, or "no pain," end of the scale showed that 18 subjects selected the 2,500-Hz carrier frequency and 9 subjects selected the 5,000-Hz carrier frequency. No subject selected bursts of the 10,000-Hz carrier frequency (Tab. 3).

Peak current to produce the muscle contractions at 50% of MVC was consistently similar among burst combinations with a specific carrier frequency. Mean milliamperage among the combined bursts (50, 70, and 90 bps grouped together) with specific carrier frequencies of 2,500, 5,000, and 10,000 Hz were significantly different (F=83.42, df=8 and 208, P<.001). Post hoc analysis revealed that combined bursts with 2,500 Hz (X [bar]=35.4, SD=8.11 mA) differed from combined bursts with 5,000 Hz (X [bar]=43.4, SD=10.44 mA) and 10,000 Hz (X [bar]=19.7, SD=2.99 mA). The combinations also differed between carrier frequencies of 5,000 and 10,000 Hz. Correlations between milliamperage used to induce torques equivalent to 50% of MVC and pain intensity ratings of subjects ranged from -.13 to .33.

Approximately 50% of the subjects reported delayed onset of muscle soreness 24 to 48 hours following at least one stimulation session. These reports were given by subjects throughout the study.

There was no single combination of the nine burst and carrier frequency combinations that was optimal for producing strong muscle contractions while perceived to be less painful than others. This finding was unexpected. Burst combinations with carrier frequencies of 2,500 and 5,000 Hz did not result in differences in pain intensity. The burst combinations with the carrier frequency of 10,000 Hz were significantly more painful than the burst combinations with a carrier frequency of either 2,500 or 5,000 Hz.

The explanation for this reversal of the pain/frequency concept may lie in part in the "jerking" sensation (ie, unfused tetanic-appearing contraction response) reported by all the subjects tested with the 10,000-Hz carrier frequency. The muscle contractions caused by this carrier frequency may also have caused increased subject apprehension. Several subjects commented that "it felt like my muscles were tearing." This apprehension caused several subjects to attempt contractions of their hamstring muscles in order to alleviate the jerking motion and pain. As the attempts by some subjects to voluntarily contract the hamstring muscles were immediately discouraged by the tester, the brief contractions were not considered contributory to the recorded "jerking" torque waveforms. Perhaps the 10,000-Hz carrier frequency with the least amount of effective current and pulse charge among the tested carrier frequencies did not adequately excite as many motor units, especially the deeper ones, thereby restricting the muscle from going into fused tetany. Perhaps the damp setting of the torque recorder used to calibrate the dynamometer and record torque smoothed twitchlike contractions to give the appearance of a jerking motion or unfused tetany. This observation of torque curves associated with an appearance of unfused muscle tetany needs further investigation.

When further examining the data of pain intensity responses to the nine burst and carrier frequency combinations, certain trends were observed. As the rate of bursts increased, the pain intensity decreased, and subjects also reported more pain as the carrier frequencies increased. The phase duration of the carrier frequency of 10,000 Hz is 50 microseconds, which is shorter than the phase duration with either 2,500 and 5,000 Hz. Others[15,39] have reasoned that subject comfort improves with very short phase durations (eg, 2,500-10,000 Hz). This concept of a relation of pain and frequency seems to apply for frequencies of continuous pulsed current but apparently not to burst and carrier frequency combinations. Our findings at burst combinations with the carrier frequency of 10,000 Hz are in disagreement with the findings of others[20] because this frequency was found to be very painful in our study. Our pain intensity scores decreased, however, as the burst rate increased from 50 to 90 bps.

As pain-conducting nerve fibers are believed to be more superficially located than motor fibers,[40,41] the stimulation by the carrier frequencies (eg, 2,500 Hz) is believed to minimize pain fiber stimulation,[1] in spite of the greatest current density being located directly beneath each stimulating electrode.[42] Researchers, using different[15,16] or similar[20] stimulus characteristics than those of this study, reported a decrease in painful sensory response as the frequency of the stimulation increased. Our finding of burst and carrier frequency combinations does not follow that reported observation. That is, as the carrier frequencies increased from 2,500 to 5,000 to 10,000 Hz in our study, subjects perceived the higher frequencies as more painful.

The pain intensity ratings reported in this study ranged considerably, with subjects marking the VAS in the upper range for the burst combinations with the 10,000-Hz carrier frequency. No subject found any burst rate accompanied by a 10,000-Hz carrier frequency to be an acceptable combination. Because of the amount of pain experienced by subjects receiving the 10,000-Hz carrier frequency, we believe the present use of this frequency in combination with bursts is questionable. Although the ratings on the VAS favor the burst combinations of the 2,500-Hz carrier frequency over the other combinations, no statistical significance was found between mean pain intensity ratings of burst combinations with either the 2,500-Hz or the 5,000-Hz carrier frequency. This statistical finding implies that the physical therapist can apply burst combinations with a carrier frequency of 2,500 or 5,000 Hz without patients perceiving any difference in pain intensity.

Table 3. Number of Subjects Selecting Combinations of Burst and Carrier
Frequencies as Most Comfortable (N=27)
                       Carrier Frequency (Hz)
Burst
Frequency (bps)(a)     2,500    5,000    10,000    Total
50                       4        2        0         6
70                       5        3        0         8
90                       9        4        0        13
Total                   18        9        0        27

The milliamperage increased as expected for carrier frequencies of 2,500 and 5,000 Hz. This increase of peak current follows the concept of strength duration, in that current amplitude increases as the pulse or phase durations decrease.[43] The results found for subject responses to the carrier frequency of 10,000 Hz disagree with the strength-duration concept because the milliamperage decreased from that required to produce torque of 50% of MVC. In general, the correlation coefficients indicated that little to no relation existed between the two variables (milliamperage and pain intensity ratings), which agrees with the findings of others.[44]

Subjects more often chose burst combinations with the 2,500-Hz carrier frequency than with any other carrier frequency. The 2,500-Hz carrier frequency had the longest phase duration among the combinations studied. This finding appears to agree with that of Baker and coworkers,[19] who reported that pulses of 300 microseconds' duration were perceived as more comfortable than those of 50 microseconds' duration. This observation is consistent when considering the amount of peak current delivered to obtain the equivalent of 50% of MVC with NMES in this study. The peak current recorded in our study was greater for the burst combinations with the 5,000-Hz carrier frequency than for either the 2,500-Hz or the 10,000-Hz carrier frequency when readings were taken of the mean milliamperage reported for these stimulus combinations. Stimulus characteristics that are comfortable at one level do not appear to guarantee comfort when applied at other levels of muscle contraction.[18] Thus, the clinician may have to try different stimulus combinations on subjects at different current training levels.

Because of subject variation on VAS ratings in this study, we recommend that patients be given a choice of carrier frequencies prior to using the "Russian" current for muscle performance enhancement. For example, as the number of subjects selecting combinations of burst frequencies (50, 70, and 90 bps) with 2,500 Hz and 90 bps with 5,000 Hz were most commonly associated with the least pain, the patient could be given a brief stimulus (10 seconds) of each. From this stimuli exposure, the patient would be able to select the combination of choice for treatment or training purposes. Delayed onset of muscle soreness emerged as the study progressed, but because subjects rated their pain immediately following each stimulation sequence, no VAS ratings covered this situation.

Conclusion

No statistically significant difference was found between combinations of burst frequencies (50, 70, and 90 bps) and carrier frequencies of 2,500 and 5,000 Hz. All burst combinations with carrier frequencies of 2,500 and 5,000 Hz differed significantly from burst combinations with the carrier frequency of 10,000 Hz. Data thus permit the conclusion that subjects be given trials of selected combinations of burst and carrier frequencies to choose the least painful stimulus combination.

References

[1] Babkin D, Timtsenko N, trans. Notes from Dr YM Kots's (USSR) lectures and laboratory periods, Canadian-Soviet exchange symposium on electrostimulation of skeletal muscles; Concordia University, Montreal, Quebec, Canada; December 6-15, 1977. [2] Halbach JW, Straus D. Comparison of Electro-Myo stimulation to isokinetic training in increasing power of the knee extensor mechanism. Journal of Orthopaedic and Sports Physical Therapy. 1980;2:20-24. [3] Kramer JF, Mendryk SW. Electrical stimulation as a strength improvement technique: a review. Journal of Orthopaedic and Sports Physical Therapy. 1982;4:91-98. [4] 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. [5] Currier DP, Mann R, Muscular strength development by electrical stimulation in healthy individuals. Phys Ther. 1983;63:915-921. [6] McMiken DG, Todd-Smith M, Thompson C. Strengthening of human quadriceps muscles by cutaneous electrical stimulation. Scand J Rehabil Med. 1983;15:25-28. [7] Owens J, Malone T. Treatment parameters of high frequency electrical stimulation as established on the Electro-Stim 180. Journal of Orthopaedic and Sports Physical Therapy. 1983;4:162-168. [8] Selkowitz DM. Improvement in isometric strength of the quadriceps femoris muscle after training with electrical stimulation. Phys Ther. 1985;65:186-196. [9] Kramer JF. Effect of electrical stimulation current frequencies on isometric knee extension torque. Phys Ther. 1987;67:31-38. [10] Nitz AJ, Dobner JJ. High intensity electrical stimulation effect on thigh musculature during immobilization for knee sprain: a case report. Phys Ther. 1987;67:219-222. [11] 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. [12] Lai HS, DeDomenico G, Strauss GR. The effect of different electro-motor stimulation training intensities on strength improvement, Australian Journal of Physiotherapy. 1988;34: 151-164. [13] Knuttgen HG. Development of muscular strength. In: Knuttgen HG, ed. Neuromuscular Mechanisms for Therapeutic and Conditioning Exercise. Baltimore, Md: University Park Press; 1976:107. [14] Kramer JF. Muscle strengthening via electrical stimulation. CRC Crit Physical Rehabilitation Medicine. 1989;1:97-133. [15] Vodovnik L, Long CP, Lippay C. Pain response to different tetanizing currents. Arch Phys Med. 1965;46:187-192. [16] Crochetiere WJ, Vodovnik L, Reswick JB. Electrical stimulation of skeletal muscle. Med Biol Eng. 1967;5:111-125. [17] Wong RA. High voltage versus low voltage electrical stimulation: force of induced muscle contraction and perceived discomfort in healthy subjects. Phys Ther. 1986;66:1209-1214. [18] Delitto A, Rose SJ. Comparative comfort of three waveforms used in electrically eliciting quadriceps femoris muscle contractions. Phys Ther. 1986;66:1704-1707. [19] Baker LL, Bowman BR, McNeal DR. Effects of waveform in comfort during neuromuscular electrical stimulation. Clin Orthop. 1988;233: 75-85 [20] Moreno-Aranda J, Seireg A. Force response to electrical stimulation of canine skeletal muscle. J Biomecb. 1981;14:595-599. [21] Myklebust BM, Robinson AJ. Instrumentation. In: Snyder-Mackler, Robinson AJ, eds. Clinical Electrophysiology: Electrotherapy and Electrophysiologic Testing. Baltimore, Md: Williams & Wilkins; 1989:27. [22] Electrotherapeutic Terminology in Physical Therapy. Alexandria, Va: Section on Clinical Electrophysiology, American Physical Therapy Association; 1990:23. [23] Kots YM; Lightfoot P, trans. Training with the method of electric tetanic stimulation of muscle by orthogonal impulses. Theory and Practice of Physical Culture. 1971;4:66. [24] Delitto A, Robinson AJ. Electrical stimulation of muscle: techniques and applications. In: Snyder-Mackler L, Robinson AJ, eds. Clinical Electrophysiology: Electrotherapy and Electrophysiologic Testing. Baltimore, Md: Williams & Wilkins; 1989:101. [25] Johnston RM, Kasper S. Compound nerve action potentials produced by signals from clinical stimulators. Phys Ther, 1986;66:85. Abstract. [26] Edwards RHT, Young A, Hosking GP, Jones DA. Human skeletal muscle function: description of tests and normal values. Clin Sci Mol Med 1977;52:283-290. [27] Currier DP. Positioning for knee strengthening exercises. Phys Ther. 1977;57:148-152. [28] Richard G, Currier DP. Back stabilization during knee strengthening exercise. Phys Ther. 1977;57:1013-1015. [29] Snyder-Mackler L, Garrett M, Roberts M. A comparison of torque generating capabilities of three different electrical stimulating currents. Journal of Orthopaedic and Sports Physical Therapy. 1989;10:297-301. [30] Brooks ME, Smith EM, Currier DP. Effect of longitudinal versus transverse electrode placement on torque production by the quadriceps femoris muscle during neuromuscular electrical stimulation. Journal of Orthopaedic and Sports Physical Therapy. 1990;11:530-534. [31] Nolan MF. Conductive differences in electrodes used with transcutaneous electrical nerve stimulation devices. Phys Ther 1991;71: 746-751. [32] Bigland-Ritchie B, Furbush F, Woods JJ. Fatigue of intermittent submaximal voluntary contractions: central and peripheral factors. J Appl Physiol. 1986;61:421-429. [33] Price DD, McGrath PA, Rafii A, et al. The validation of visual analogue scales as ratio scale measures for chronic and experimental pain. Pain. 1983;17:45-56. [34] Scott J, Huskisson EC. Accuracy of subjective measurements made with or without previous scores: an important source of error in serial measurements of subjective states. Ann Rheum Dis. 1979;38:558-559. [35] Ohnhaus E, Adler R. Methodological problems in the measurement of pain: a comparison between the verbal rating scale and the visual analog scale, Pain. 1975;1:379-384. [36] Revill S, Robinson J, Rosen M, et al. The reliability of a linear analog for evaluating pain. Anaesthesia. 1976;31:1191-1198. [37] Downie WW, Leatham PA, Rhind VM, et al. Studies with pain rating scales. Ann Rheum Dis. 1978;37:378-381. [38] Safrit MJ, ed. Reliability Theory. Washington, DC: American Alliance for Health, Physical Education, and Recreation; 1976;17-24. [39] Lloyd T, DeDomencio G, Strauss GR, et al. A review of the use of electromotor stimulation in human muscles. Australian Journal of Physiotherapy. 1986;32:18-30. [40] Melzak R. Prolonged relief of pain by brief, intense transcutaneous electrical nerve stimulation. Pain. 1975;1:357-373. [41] Quirion-DeGirardi CQ, Seaborne D, Goulet FS. The analgesic effect of high voltage galvanic stimulation combined with ultrasound in the treatment of low back pain. Physiotherapy Canada. 1984;36:327-333. [42] Cook TM. Instrumentation. In: Nelson RM, Currier DP, eds. Clinical Electrotherapy. East Norwalk, Conn: Appleton & Lange; 1987:22. [43] Howson DC. Peripheral neural excitability: implications for transcutaneous electrical nerve stimulation. Phys Ther. 1978;58:1467-1473. [44] Gracanin F, Trnkoczy A. Optimal stimulus parameters for minimum pain in the chronic stimulation of innervated muscle. Arch Phys Med Rebabil. 1975;56:243-249.

The study by Rooney et al is another attempt to search for the combination of current characteristics that will bring about high contractile forces while simultaneously affording a tolerable experience for the subject. Their approach of manipulating current characteristics (ie, burst rates and carrier frequencies) is analogous to the approaches of other studies[1-4] (including one by this author), and their conclusion is very much the same as those of other studies, namely, that individual preferences exist for various current forms. In commenting on this study, I would like to specifically focus on three areas: the authors' pain scaling methods, their torque measurements, and their explanation of electrically associated discomfort.

It is well established that the phenomenon of pain is multidimensional and includes distinctly different dimensions commonly labeled as "intensity" and "unpleasantness."[5] In addition to the obvious intensity directly resulting from stimulation of afferent fibers, other factors associated with being electrically stimulated will certainly invoke affective components of pain, especially to the electrically naive subject. It is the ability to elicit both intensity and affective components of pain that makes using electrical stimulation an attractive methodology in experimental pain studies.[6]

The authors' pain scaling method is confusing in that they imply throughout the manuscript that they intended to measure pain intensity, yet they label the extremes of their scale with adjectives that are not the best descriptors of pain intensity (ie, "no pain" and "pain as bad as it could be"). The authors ignore any mention of affective pain components until the discussion, in which they attempt to explain their only significant finding (discomfort associated with the highest carrier frequency) using the term "apprehension," a classical affective pain descriptor. If apprehension was indeed the reason subjects felt increased discomfort with the highest carrier frequency, then this brings into serious question the authors' contention that their scale was measuring pain intensity. Perhaps with more appropriate pain intensity descriptors as well as the addition of an affective pain scale, the authors could have more precisely defined the discomfort associated with the highest carrier frequency.

I would be curious to know the authors' rationale for using a damp of 3 for the isokinetic torque measures, especially in light of the problems such a high damp setting will introduce to a torque trace.[7,8] It is well established that with a known load applied to a lever arm, increasing a damp setting decreases the amplitude of torque tracings but the amount of decrease is dependent on the size of the load and thus virtually impossible to predict (see Figure).[9] Introducing such an artifact to the torque trace raises questions as to the accuracy of the authors' claims of electrically eliciting quadriceps femoris muscle contractions to 50% of a maximal volitional contraction. The issue of excessively high damp also raises a question concerning the authors' claim that torque fluctuations occurred only at the highest frequency. Higher damp settings result in excessive capacitance in the torque signal that will cause the pen tracing to be less responsive to any fluctuations in loads applied to the lever arm, such as those the authors propose are happening at the highest carrier frequency. My question is not whether such fluctuations are indeed occurring at the highest carrier frequency, but, How can the authors be sure that substantial fluctuations are not occurring at lower carrier frequencies and that they are not able to illustrate such fluctuations because the torque trace is excessively damped?

The authors strongly imply that stimulation of nociceptors is the sole factor in discomfort during electrically elicited muscle contractions. I would suggest that an individual's ability to tolerate an electrically elicited muscle contraction may be more complex than the authors' explanation, which implicates the stimulation of nociceptors that are concomitantly stimulated with alpha motoneurons. Issues such as an individual's coping style, an individual's perceived control during the event, and other cognitive-behavioral aspects related to pain tolerance are known to have an effect on pain reports during an aversive event.[10-14]

As clinicians, we would all like to make electrical stimulation as tolerable to patients as possible. Advancing technology now allows electrical currents to be delivered to patients in a variety of fashions, leading to studies such as those of Rooney et al and others,[1-4] which suggest that changing current characteristics may improve patient comfort. The results of these studies become difficult to apply to clinical situations in that they are all conducted on subjects for whom electrical stimulation is not indicated. Thus, it remains to be seen in applied situations exactly how much more tolerable electrical stimulation can become using such strategies as changing current characteristics.

References

[1] Delitto A, Rose SJ. Comparative comfort of three waveforms used in electrically eliciting quadriceps femoris muscle contractions. Phys Ther. 1986;66:1704-1707. [2] Reisman MA. A comparison of electrical stimulators in eliciting muscle contractions. Phys Ther 1984;64:751. Abstract. [3] Grimby G, Wigerstadt-Lossing I. Comparison of high- and low-frequency muscle stimulators. Arch Phys Med Rehabil 1989;70:835-838. [4] Baker LL, Bowman BR, McNeal DR. Effects of waveform on comfort during neuromuscular electrical stimulation. Clin Orthop. 1988;233: 75-85. [5] Melzack R, Torgerson WS. On the language of pain. Anesthesiology. 1971;34:50-59 [6] Rollman GB, Harris G. The detectability, discriminability and perceived magnitude of painful electric shock. Perception and Psychophysics. 1987;42:257-268. [7] Rothstein JM, Lamb RL, Mayhew TP. Clinical uses of isokinetic measurements: critical issues. Phys Ther. 1987;67:1840-1844. [8] Mayhew TP, Rothstein JM. Measurement of muscle performance with instruments. In: Rothstein JM, ed. Measurements in Physical Therapy. New York, NY: Churchill Livingstone Inc; 1985;7:57-102. [9] Sinacore DR, Rothstein JM, Delitto A, Rose SJ. Effect of damp on isokinetic measurements. Phys Ther. 1983;64:1248-1250. [10] Averill JR. Personal control over aversive stimuli and its relationship to stress. Psychol Bull. 1973;80:286-303. [11] Miller SM. Controllability and human stress: method, evidence and theory. Behav Res Ther 1979;17:287-304. [12] Thompson S. Will it hurt less if I can control it? A complex answer to a simple question. Psychol Bull. 1981;90:89-101. [13] Miller SM. Monitoring and blunting: validation of a questionnaire to assess styles of information seeking under threat. J Pers Soc Psychol 1987;52:345-353. [14] Tan S. Cognitive and cognitive-behavioral methods for pain control: a selective review. Pain, 1982;12:201-228.

Author Response

We would like to thank Dr Delitto for his review of our study. We agree that pain is a multidimensional phenomenon. Although it may be true that pain has both sensory and affective components,[1,2] it is not at all clear that pain measures incorporating both dimensions (especially the affective domain) are necessarily sensitive to varying levels of pain stimulation in the laboratory.[3] The visual analog scale (VAS), on the other hand, has a strong track record of sensitivity to variations in pain intensity, is highly reliable,[4-6] and correlates strongly with other more global pain measures.[7] For making multiple pain assessments over a relatively brief period of time, the VAS format is ideal because it is difficult for subjects to remember the pain intensity rating of a previous trial. In our opinion, using an affective scale with words such as "tiring," "sickening," "fearful," and "punishing" would not have added significantly to our findings.

Dr Delitto also expressed concern about our use of the term "apprehension" in our discussion. In our discussion, we attempted to indicate that we were not using the term as a classical affective pain descriptor, but rather were conveying our interpretation of what the subjects were describing (eg, "My muscles were tearing").

The pen-writer-type galvanometer is used in the Cybex[R] dynamometer system.(*) Overshoot or deflection of the pen upon sudden inertial impact and frictional forces is undesirable,[8] and damp is particularly important in galvanometer-like recorders to overcome oscillations when a load is suddenly applied to the recorder.[9] The method cited by Dr Delitto[10] performed isokinetic measurements at speeds of 30 [degrees], 75 [degrees], and 180 [degrees]/s. Our study used a 0 [degrees]/s speed (isometric) at a fixed 60-degree knee and hip angle. Experience with the damp may be somewhat different with the two studies. Overshoot does occur at 0 [degrees]/s, as it does at speed rates of 30 [degrees] to 180 [degrees]/s, but the torque tracing levels off immediately to provide a constant reading with isometric torque measures. There is no displacement of the torque curve to the right because there is no movement of the torque arm through a range of motion. in the referenced study,[10] the mean decreases in amplitude of torque tracings were about 1.6% for the 0 damp setting and 0.8% for the damp setting of 3 (that is, less error was reported with the latter setting). In our laboratory, the changes in amplitude of isometric torque tracings averages 1% or less.

Dr Delitto mentions that coping style and other cognitive-behavioral aspects are related to a subject's ability to tolerate pain induced by electrically elicited muscle contraction. Although such an observation is true, when looking across subjects with a wide variety of coping styles, it is likely that data collected from such observations would increase the "background noise" and unnecessarily obscure the significant findings of this study. By no means should we reject the significant findings of this study because variation in coping styles may have contributed to the effect. (*) Cybex, Div of Lumex Inc, 2100 Smithtown Ave, Ronkonkoma, NY 11779.

References

[1] Melzack R. The McGill pain questionnaire: major properties and scoring methods. Pain. 1975;1:277-299. [2] Melzack R. The short form of the McGill pain questionnaire. Pain. 1987;30:191-197. [3] Bruehl S, Carlson CR, McCubbin JA. The relationship between pain sensitivity and blood pressure in normotensives. Pain. 1992; 48:463-467. [4] Scott J, Huskisson EC. Accuracy of subjective measurements made with or without previous scores: an important source of error in serial measurements of subjective states. Ann Rheum Dis. 1979;38:558-559. [5] Ohnhaus E, Adler R. Methodological problems in the measurement of pain: a comparison between the verbal rating scale and the visual analog scale, Pain. 1975;1:379-384. [6] Revill S, Robinson J, Rosen M, et al. The reliability of a linear analog for evaluating pain. Anaesthesia. 1976;31:1191-1198. [7] Downie WW, Leatham PA, Rhind VM, et al. Studies with pain rating scales. Ann Rheum Dis. 1978;37:378-381. [8] Suckling EE. Bioelectricity. New York, NY: McGraw-Hill Inc; 1961:46-47, 89-91. [9] Strong P. Biophysical Measurements. Beaverton, Ore: Tektronix Inc; 1973:403. [10] Sinacore DR, Rothstein JM, Delitto A, Rose SJ. Effect of damp on isokinetic measurements. Phys Ther. 1983;63:1248-1250.