A study of discomfort with electrical stimulation.

Transcutaneous neuromuscular electrical stimulation (NMES) has many therapeutic applications, including electrically induced analgesia (usually referred to as transcutaneous electrical nerve stimulation), diagnostic uses (eg, nerve conduction studies), and the rehabilitation of muscle weakness. It is a well-documented method of improving muscle performance in both healthy and weak muscle.[1,2] Possible mechanisms of NMES-induced improvement in muscle performance include (1) identical mechanisms that occur with volitional exercise training, including well-established principles emphasized under the guise of "functional overload," and (2) selective recruitment of type 11 muscle fibers, which more effectively targets fast-twitch motor units and conceivably offers more efficient strength gains.[2] With either explanation, eliciting a muscle contraction via electrical stimulation may necessitate that patients tolerate what can be an uncomfortable procedure.

Selkowitz[3] has shown isometric torque gains resulting from NMES training regimens to be positively and linearly correlated with electrically elicited muscle contractile forces expressed as a percentage of maximal volitional contractile forces. Selkowitz's findings support the biological principles of functional overload and suggest that in order to obtain maximal benefit from NMES regimens, current intensities must be sufficient to elicit relatively high contractile forces.

Investigators have yet to determine the minimally sufficient dosage of current that will elicit contractile forces capable of successfully training muscle, but past studies[3-6] have shown successful muscle strengthening after NMES training regimens ranging from 33% to 91% of maximal volitional contraction (MVC).[3-6] Electrically eliciting muscular force within the upper limits of this range requires high current intensities, with accompanying subject pain levels that may approach an intolerable level. Thus, as with most clinical interventions, there is a therapeutic goal (muscle strengthening) and a cost (pain to the subject) with NMES. An emphasis on muscle strengthening at the expense of pain raises obvious ethical concerns. Alternatively, ignoring the therapeutic goal of muscle strengthening can result in current intensities that are therapeutically ineffective.

Subject discomfort is often the limiting factor during NMES, and it is difficult to predict clinically. At this time, little is known about how subjects define the discomfort that accompanies NMES. In the past, the discomfort associated with NMES has been measured using visual analog scales,[7,8] categorical scales,[9] or paired-choice techniques.[10] Three major limitations are evident from the previous studies cited. First, the major focus has been the subject's preference for various current forms, clearly illustrating that a subject's tolerance of NMES was investigated as though it was solely a function of the stimulus (eg, current characteristics) and the interface between the electrode and the afferent receptors. Second, past work has failed to consider individual differences as moderators of tolerance for NMES.[7-10] Finally, the discomfort associated with NMES was measured as though pain associated with NMES was unidimensional. This approach ignores a substantial body of literature that describes pain as a multidimensional experience. We examined each of these issues in this investigation.

Previous Research Has Been

Overly Focused on Afferent

Stimulation

The most recent advances attempting to make NMES more palatable to subjects have focused on the current characteristics' role in pain production.[3-6] Although direct stimulation of afferent nerves results in some of the discomfort associated with NMES, there may be other factors (eg, cognitive, moderating) that can add to this discomfort. For example, electrically elicited muscle contractions are involuntary and require a subject to tolerate what can be high contractile forces. These high contractile forces can indirectly provoke increased pain, apprehension, or fear, resulting in added discomfort associated with NMES. Thus, we believe that the feelings associated with the muscular contraction and the effect of direct stimulation of afferent nociceptors combine to produce the discomfort that people feel. This would suggest that there is a distinction between the pain from the physical stimulus and the discomfort from the muscular contraction. In this study, we separated the pain response attributable to muscle contraction from that attributable to afferent stimulation alone.

Cognitive Moderation of Pain

The traditional investigative approach treats the sources of NMES-associated discomfort as though the subject is a "hard-wired" system and assumes that a particular set of current characteristics (eg, waveform, pulse rate, frequency) will be best tolerated by all people.[7-10] This approach is mechanical in concept and ignores well-established cognitive and behavioral components, such as the concept of control,[11-13] that may be associated with an aversive event such as NMES. An investigative approach that includes study of any cognitive-behavioral component associated with the discomfort felt with NMES could prove useful, especially in light of (1) individual coping strategies that have been identified[14] and (2) applied evidence indicating that coping with an aversive event is moderated by coping strategies.[15]

Of the numerous coping strategies investigated, those involving the concept of control are the most elaborated. Thompson defines control as "the belief that one has at one's disposal a response that can influence the aversiveness of an event."[13](p89) She further defines control using a fourfold typology consisting of (1) behavioral control, (2) cognitive control, (3) information, and (4) retrospective control. Behavioral control is the belief that one has a behavioral response available that can affect the aversiveness of an event. Cognitive control is the belief that one has a cognitive strategy available that can affect the aversiveness of an event. information, although not necessarily a form of control, refers to any communication delivered to a person who is the potential recipient of an aversive event and is included because such information can "engender" feelings of control. Retrospective control refers to feelings or beliefs about an aversive event that occur subsequent to the event.

Information is particularly relevant to the study of NMES because much of the aversiveness is presumed to arise from its novel features (eg, the direct stimulation of afferent nerves). Information, however, can increase stress that would not ordinarily be present during an aversive event. Thompson[13] uses an example of the city of Los Angeles providing information to citizens that was designed to decrease stress in elevator riders. The information included the remoteness of the possibility than elevator car would drop uncontrollably or that an elevator car would run out of air. The result of providing such information was to increase sum in a number of people, probably because fears normally not present were elicited by the notices.

Perhaps the issue as to whether stress is relieved or exacerbated by providing information remains clouded because many experiments fail to note individual differences among subjects with respect to whether they want information about a stressful event. Miller[14] contends that when people are confronted with an aversive event, they differ in how they handle information about the event. Consequently she has developed a scale that characterizes subjects as either "monitors" (information seekers) or "blunters" (information avoiders) based on self-reported preferences for information or distraction as subjects imagine themselves involved in a variety of stressful situations.[14]

Miller[14] contends that, when faced with a potentially aversive event, monitors actively seek information, presumably to reduce uncertainty concerning the event. Blunters avoid information-seeking strategies, preferring to distract themselves during aversive situations. Miller has demonstrated that monitors and blunters exhibit reliable preferences in seeking information during potentially stressful or aversive events. For example, when faced which a potentially aversive event such as the prospect of electric shock, high-monitors/low-blunters hereafter referred to as "monitors") spend more time using an information-seeking strategy in comparison with high-blunters/low-monitors (hereafter referred to as "blunters"). The blunters prefer listening to distracting music. In addition, blunters experience less anxiety than monitors when faced with threatening situations.

The question of providing subjects with information about an aversive event becomes more complex when die event is a medical procedure requiring informed consent. There is an ethical consideration that subjects make an informed decision regarding their participation in die event. Yet, detailed procedural information may induce stress by providing information that would not ordinarily be thought of by the subject, similar to the elevator story described previously.

Subjects must be given enough information to make an informed choice to participate in a procedure. Commonly, however, information concerning the medical procedure exceeds that necessary for customary informed consent compliance. With NMES, for example, specific information concerning what the stimulus feels like, where the subject will feel it, and sensations resulting from the muscle contraction are commonly thought to be beneficial in raising a subject's tolerance level to the stimulus. The argument could be made that for some individuals who prefer distraction (eg, blunters), these procedures may actually decrease tolerance to the procedure.

In this study, we examined pain reports in response to NMES as moderated by the monitoring-blunting distinction. Given the information-rich nature of the typical informed consent procedure, we expected NMES to be better tolerated by monitors than by blunters.

Multidimensional Nature

of Pain

Since the early 1970s, at least two dimensions of pain have been recognized. Melzack and Torgerson[16] produced a multidimensional pain descriptor scale with three major classes: sensory quality, affective quality, and an overall pain evaluation. This scale has been used in both clinical and experimental pain investigations. The latter studies included electrical stimulation in the methodologies as a technique for administering experimental pain. Thus, it seems reasonable to assume that the discomfort associated with therapeutic NMES can be labeled as painful and that pain has a sensory component (intensity), as well as an affective component (unpleasantness).

The measurement of intensity and unpleasantness of pain stimuli has been refined further, as evidenced by recent psychophysical studies of pain by Gracely and colleagues (for a review, see Gracely[17]). In their investigations involving acute clinical pain, Gracely and colleagues[17,18] have included the independent assessment of the intensity and unpleasantness of painful stimuli using ratio-scaling techniques. Clinical examples used by Gracely and colleagues,19,20 such as documenting the analgesic effects of narcotics during specific dental techniques, closely parallel the technique of NMES. Both techniques represent potentially aversive situations that a patient must endure, and the aversiveness can be argued to contain both intensity and unpleasantness components.[21]

Intensity and unpleasantness are distinctly different dimensions of pain. Clarifying the aversiveness of NMES-associated pain will not only lead to a better understanding of the discomfort associated with NMES, but perhaps may suggest better strategies for alleviating subject discomfort during electrically elicited muscle contractions.

In addition, measuring intensity and unpleasantness using the technique of magnitude estimation should further elucidate the pain associated with NMES. Magnitude estimation offers two primary advantages over categorical or ordinal ratings that have been used in the past with NMES comfort studies.[7-10] First, magnitude estimation produces ratio scales, which allow a clearer estimation of functional relations. Second, magnitude estimation is assumed to produce more reliable measures. The first advantage hinges on the assumption that subjects can render judgments that conform to ratio-level expectation, thus providing highly sensitive measures of stimulation intensity. Lodge[2]2 reports extensive evidence that such judgments are easy to produce, and other research using magnitude estimation verifies the ratio character of such judgments.[17] The presumed high reliability of magnitude estimates is indicated by the greater proportion of variance that such measures can account for in other outcome variables compared with categorical scales and by the high correlations (>.90) among magnitude estimates obtained using different response modalities (eg, lines and numbers).[22]

Study

There are no reports describing the discomfort associated with NMES that take into account both cognitive behavioral influences and the multidimensionality of pain. The purposes of this study were (1) to discern whether a behavioral coping style has any relation to how a subject describes the discomfort associated with NMES, (2) to investigate and contrast the magnitude of these components during NMES resulting in electrically elicited muscle contractile forces and NMES resulting in only direct stimulation of afferent fibers, and (3) to characterize the discomfort associated with NMES according to two well-described components of pain (intensity and unpleasantness).

Methods

A general overview of the methodology used in this investigation is provided in Table 1. The study took part in two phases. First, coping styles of a group of individuals (N = 70) were determined using the Miller Behavioral Styles Scale (MBSS) questionnaire.[14] Informed consent was obtained before the questionaire was administered. This scale asks the subjects to imagine themselves in four separate stress-provoking situations. Statements following each situation represent different ways in which a subject might deal with the situation. One half of the choices indicate an information-seeking strategy, and the other half of the choices indicate an information-avoiding strategy. Subjects are asked to indicate the statements that best describe the strategies that they would use in dealing with the situation. Two separate scores are derived from the MBSS, one representing a total monitoring score and the other representing a total blunting score. An example of one portion of the MBSS is illustrated in Figure 1.

Table 1. Summary of Methodology
Step   Procedure
1      Administer and score Miller Behavioral Styles Scale
         questionnaire
2      Select monitors and blunters; dismiss from the study
         those subjects who had no clear coping style
3      Randomly select current type (muscle current/afferent
         current, afferent current only)
4      Determine MVIT(a) (with muscle current/afferent current
         only)
5      Determine maximal current tolerated; record any
         electrically elicited torque production with muscle
         current/afferent current
6      Calculate 45%, 60%, 75%, and 90% of maximal current
         tolerated
7      Randomly select pain descriptor to be judged (intensity,
         unpleasantness)
8      Administer 20-mA DC reference current
9      Administer current levels randomly; record magnitude
         estimate for pain descriptor (intensity or
         unpleasantness) for each current
         level; have subject choose most appropriate adjective
         from the appropriate pain descriptor category in Table
         2 from each current level
10     Repeat steps 8 and 9 for the remaining pain descriptor
11     Wait at least 72 hours, then repeat steps 4 through 10
       with the remaining current type
(a) MVIT = maximal volitional isometric torque production.

After the scores of the MBSS were tallied and assessed, a subgroup (n=30) was identified whose coping strategies were consistently either monitoring (n= 15) or blunting (n=15). The remaining 40 subjects were dismissed from the study.

Phase 2 required both monitors and blunters to be subjected to two techniques of delivering NMES on two separate occasions. The examiner (AD) remained blinded to which group (monitors, blunters) each subject belonged. One technique (muscle current/afferent current, or current type A) resulted in a forceful muscle contraction. The other technique (afferent current only, or current type B), delivered with the same current generator(*) at similar current amplitudes and characteristics (2,500 Hz triangular carrier frequency, 75 pulses per second), resulted in predominantly afferent stimulation. The major difference between the current techniques was electrode placement. In addition, muscle torque measurements were obtained during the muscle current/afferent current technique.

Muscle Current Technique

Each subject was seated with the hips at 90 degrees and the tested knee in 65 degrees of flexion at a dynamometer(dagger) set in an isometric mode. Hips were stabilized with a stabilization bar. Two self-adhesive electrodes,(double dagger) 7.62 x 12.7 cm, were placed over the quadriceps femoris muscle group, one over the muscle belly of the vastus medialis oblique muscle and the other over the most proximal area of the vastus lateralis muscle. The shank was then attached to the lever arm of the dynamometer, the knee axis of rotation aligned with the axis of the dynamometer, and the pelvis stabilized with a padded bar.

An isometric MVC of the quadriceps femoris muscle was obtained and the resultant isometric knee extension torque was recorded. Contraction duration was approximately 3 seconds. To account for any learning effect, this procedure was repeated at least twice. If the first or second contraction resulted in the highest torque reading of the three repetitions, then this level was labeled the maximal volitional torque recording. If the third contraction resulted in a maximal torque reading, of additional measurements were taken until a decrease in torque reading was obtained. The highest of any subsequent torque reading was labeled the maximal volitional torque recording. There was a 30-second rest period between contractions.

Afferent Current Technique

Each subject was seated in the position described previously. Two self-adhesive electrodes measuring 7.62 x 10.16 cm were affixed to the skin of the subject's knee on the lateral and medial joint fine adjacent to the patella in areas devoid of muscle.

The remainder of the methodology was the same for both the muscle current and afferent current procedures. First, unpleasantness and intensity of pain were examined. The subject was shown a two-column adjective list (Tab. 2) and was told that the words on die left were those ad@ commonly used to describe die intensity or sensory components of pain and that the words on the right were those adjectives used to describe the unpleasantness or affective components of pain. The subject was then told that one of the major purposes of this study was to assess how intense and unpleasant NMES was and that it was important that the subject make the distinction between the two pain components. To emphasize this point, the examiner explained that a painful stimulus could be extremely weak in terms of intensity yet excruciating in terms of unpleasantness. In addition, the opposite extreme was pointed out (a stimulus being bearable but extremely intense). Finally, the subject was told that a stimulus could be equally intense and unpleasant.

Table 2. Descriptors Used for
Distinguishing Intensity and
Unpleasantness Components of Painful
Stimuli
Sensory                      Affective
Extremely intense            excruciating
Very intense                 intolerable
Very strong                  unbearable
intense                      agonizing
Strong                       horrible
Slightly intense             dreadful
Barely strong                frightful
Moderate                     awful
Slightly moderate            miserable
Very moderate                oppressive
Mild                         distressing
Very mild                    uncomfortable
Weak                         unpleasant
Very weak                    distracting
Extremely weak               bearable

Next magnitude estimation was explained to the subject. We decided to use magnitude estimation given the purported advantages of ratio scaling as compared with categorical techniques (eg, visual analog scales) in quantifying sensation.[23] The subject was told that a reference stimulus (20-mA DC current) would arbitrarily be labeled 100 in terms of intensity or unpleasantness and that for any subsequent stimulus the subject would be asked for a number that best compared the magnitude of the stimulus (either intensity or unpleasantness) with the reference stimulus. The subject was told that the numbers that were chosen could go up to whatever number he or she thought best described the intensity or unpleasantness of the stimulus in comparison with the reference current.

Procedural events during NMES were then explained to the subject, with emphasis on choice of electrodes, electrode placements, and reasons for the positional setup. Expected sensations were described to die subject. These sensations included tingling or buzzing under the pads. The involuntary contractile force that accompanies NMES was also described in detail including the feeling of loss of control that accompanies involuntary muscle contraction. In subjects were assured by the examiner that no harm would come to them, and they were informed about such details as the stabilizing bar being necessary to prevent injury from the involuntary contraction. Subjects were further assured that any discomfort felt during the stimulation would be short-lived and would cease with discontinuation of the current and thus was not a result of any tissue damage.

Next, a maximal tolerable current level was determined. Current amplitude was increased in increments of 10 mA and was administered for 5-second durations with a 2-second ramp time. For stimulation that did not cause a muscle contraction, there was a 1.5-minute rest between stimuli For stimulation that caused a muscle contraction, there was a 3-minute rest between stimuli. Longer rest periods after stimuli causing muscle contractions were necessary for the muscle to recover from fatigue. Once the maximally tolerable current level of each current type was determined, 45%, 60%, 75%, and 90% levels of tolerance were established.

A variable-impedance box was developed to mimic each subject's skin impedance. Skin impedance was calculated through a feature on the stimulator being used, which the primary author (AD) had previously tested and verified. The impedance box showed a dose representing the four levels of current to be preset before being administered to die subject, thus eliminating the need to adjust the intensity during the stimulus on which the subject was making a judgment. Using magnitude estimation, each subject then assessed the intensity and unpleasantness of the four levels of each type of current stimulus in the following manner. First, the current type was determined randomly. Second, whether the subject would judge intensity or unpleasantness was determined randomly. Third, the order of current level was determined randomly. Finally, another set of randomly determined current levels was established for the remaining pain descriptor. The procedure was then repeated with the remaining current type.

A reference current (20-mA DC) preceded each group of intensity or unpleasantness magnitude estimates. After each level of current was administered, each subject first gave a number in comparison with the reference current (ie, 100) and then indicated an adjective from the checklist (Tab. 2) that described the sensation resulting from the current.

A summary of the variables of interest specific to the electrical stimulation techniques, along with their definitions, is provided, as follows:

1. MVIT - volitional isometric

torque production. 2. EET - electrically elicited torque. 3. Percent - EET/MVITx100, or the

percentage of maximal volitional

torque elicited using NMES. 4. Maximal current (type A) - the

maximum current (root mean

square [RMS] current measured in

milliamperes) thee subject was able

to tolerate during the muscle contraction

technique. 5. Maximal current (type B) - the

maximum current (RMS current

measured in milliamperes) the

subject was able to tolerate during

the afferent stimulation technique.

If a difference between groups existed between MVIT and EET, this finding was interpreted as reflecting differences in quadriceps femoris muscle torque production and would not have adversely affected the interpretation of any result that was attributed to the between-groups factor. It was expected, however, that percent (at all levels) and maximal current levels for both current types would not differ across the monitor and the blunter groups. If a difference existed in these variables, this difference could confound any result attributed to thee coping-styles variable. For example, if there were between-group differences in the amount of current administered, then the results could be argued to be reflective of sensory differences between groups as opposed to coping styles.

Data Analysis

Determining coping-style groups.

Median monitoring and blunting scores for the total sample (N = 70) were determined. Monitors and blunters groups were established using the following decision rules: (1) Subjects whose monitoring score was greater than the sample's median monitoring score and whose blunting score was less than the sample's median blunting score were placed in the monitors group, and (2) subjects whose blunting score was greater than the sample's median blunting score and whose monitoring score was less than the sample's median monitoring score were placed in the blunters group.

Preliminary data analysis. Preliminary data analysis consisted of assessing differences that may have existed between the monitors and blunters groups that would confound testing of the major hypotheses. These differences included differences in demographic and electrical stimulation variables described previously. These variables were compared by group with independent t tests.

The MBSS was scored by tallying the monitoring and blunting statements that were checked by the subjects on the MBSS and creating monitoring and blunting variables. We then correlated the monitoring and blunting scores to determine whether the coping-style dimensions were orthogonal (ie, correlation = 0). For descriptive purposes, typical pain descriptors used by each group for each current level were calculated by converting each descriptor to a number (1-15) according to its place in the list in Table 2. The median was then calculated for both intensity and unpleasantness by group and current level (Tab. 3).

Table 3. Median(a) Pain Descriptors for Each Current Level by
         Coping-Style Group
Current level      Group      Intensity          Unpleasantness
Muscle current
  60%              blunter    barely strong      uncomfortable
                   monitor    strong             uncomfortable
  75%              blunter    strong             distressing
                   monitor    intense            distressing
  90%              blunter    very strong        unbearable
                   monitor    very strong        miserable
Afferent current
  60%              blunter    weak               miserable
                   monitor    weak               miserable
  75%              blunter    slightly intense   unpleasant
                   monitor    strong             distressing
  90%              blunter    intense            miserable
                   monitor    intense            oppressive
(a) Numbers were assigned to each pain description in Table 2
    in ascending order from bottom to top of each list. Medians
    were then calculated for each level within each group.

Major data analysis. Intensity and unpleasantness magnitude estimates for the 60%, 75%, and 90% levels of both current forms were normalized using the magnitude estimates from the 45% current level as the denominator. Normalized magnitude estimates were analyzed using a 2x2x2x3 (coping style x current type x pain descriptor x current level) analysis of variance (ANOVA), with the last three factors treated as repeated measures. Any significant interactions were assessed further using simple-effects effect analyses. For all simple-effects analyses, aggregate error terms were calculated.[24] All statistical analyses were conducted using a commercially available computer program.[section][25]

Results

The Pearson product-moment correlation (r) between monitoring and blunting was - .207 (P = .095). The nonsignificant correlation indicates that the total monitoring and total blunting scores were orthogonal. This finding has also been documented in other studies of the MBSS instrument and indicates that the MBSS is tapping separate levels of the construct behavioral styles.[14]

The median monitoring score was 9.0, and the median blunting score was 3.0, again similar to what has been found in the past with this scale.[14] Further analysis resulted in 15 subjects (4 male, 11 female) being assigned to the monitors group and 15 subjects (6 male, 9 female) to the blunters group.

Demographic information concerning the subjects in these two groups is presented in Table 4. There were no differences in height (t = - 1.024, df = 28, P = .315) or weight (t = - .583, df = 28, P = .565) between groups. The pertinent electrical stimulation variables (percent at all levels, maximal current levels with both current types) did not differ across groups, with the largest t value equaling - 1.992 (df = 28, P = .057). A summary of the descriptive statistics of all variables by group and sex is presented in Table 5.

Table 4.  Demographic Information by Coping-Style Group
                   Male Subjects   Females Subjects   Combined
Monitors
  N                    4              11                15
  Age (y)
    X                 25              25                26
    SD                  1.4            4.5               3.9
    Range             24-27           21-35             21-35
  Height (in(a))
    X                 70              66                67
    SD                 1.5             2.3               2.9
    Range             69-72           63-69             63-72
  Weight (lb(b))
    X                181             139               150
    SD                26              18                28
    Range            165-220         115-170           115-220
Blunters
  N                    6               9                15
  Age (y)
    X                 25              28                27
    SD                 2.2             7.9               6.3
    Range             23-a            22-44             22-44
  Height (in)
    X                 71              66                68
    SD                 1.5             2.9               3.5
    Range             69-72           61-70             61-72
  Weight (lb)
    X                187             136               156
    SD                20              29                36
    Range            160-210         106-205           106-210
(a) 1 in = 2.54 cm.
(b) 1 lb = 0.4536 kg.
Table 5. Results of Electrical Stimulation(a) by Coping-Style
         Group
                                Male       Female     Combined
                                Subjects   Subjects
Monitors
  N                                4         11          15
  MVIT (ft.lb(b))
    X                            173        132         143
    SD                            26         22          29
  EET (ft.lb)
    X                            124         85          96
    SD                            37         16          28
  Percent
    X                             71         65          67
    SD                            14         10          11
  Maximal current (type A) (mA)
    X                             54         41          44
    SD                            15          7          11
  Maximal current (type B) (mA)
    X                             47         34          37
    SD                            12          6          10
Blunters
  N                                6          9          15
  MVIT (ft.lb)
    X                            223        124          163
    SD                            23         25           55
  EET (ft.lb)
    X                            152         87          113
    SD                            47         28           48
  Percent
    X                             69         70           69
    SD                            20         15           16
  Maximal current (type A) (mA)
    X                             63         46           53
    SD                            12         11           14
  Maximal current (type B) (mA)
    X                             52         32           40
    SD                             7          5           11
(a) MVIT = maximal volitional isometric torque production;
EET = electrically elicited torque; percent = EET/MVIT x 100,
or percentage of maximal volitional torque elicited; maximal
current (type A) = maximum current subject able to tolerate
during muscle contraction technique; maximal current (type B)
= maximum current subject able to tolerate during afferent
stimulation technique.
(b) 1 ft.lb = 1.356 N.m.

The results of the four-factor ANOVA showed a significant linear main effect of level of current (F = 62; df = 1,28; P<.05). There were no other main effects. Additionally, there was a significant three-way interaction of coping style x current type x pain descriptor (F = 8.0; df = 1,28; P = .008). Finally, there was significant four-way interaction with coping style x current type x pain descriptor x current level (F = 6.6; df = 1,28; P = .016). The results of the four-way interaction are illustrated in Figures 2 through 4.

Next, simple-effects analyses were conducted. First, the coping style x pain descriptor x current type interaction was assessed at each of the three levels of current. This analysis was chosen to determine the effect of the increasing force of electrically elicited muscle contractions on pain reports. With increasing current levels, both electrode placements were expected to provide similar afferent input, but the muscle current was expected to provide additional sensory input that contributed uniquely to the subjects perception of NMES.

At the 90% level of tolerance, this analysis revealed a significant three-way interaction (F = 9.68; df = 1,28; P = .004). This result is illustrated in Figure 4. Blunters found the afferent current to be predominantly intense, whereas monitors found the same current to be predominantly unpleasant. The opposite result occurred with, the muscle current. The same analyses were carried out for the 75% and 60% levels. At the 75% level, the three-way interaction was significant (F = 5.22; df = 1,28; P = .030) and was interpreted in a similar manner as for the 90% level. At the 60% level of current, however, the three-way inter-action was not significant (F = 1.36; df = 1,28; P =.254). At this current level, blunters and monitors found both current types to be equal in intensity and unpleasantness. The results for the 60% and 75% levels are illustrated in Figures 2 and 3.

A significant coping style x pain descriptor x current type interaction was not present at the lowest current level (ie, 60%), at which electrically elicited muscle contractile force was minimal (<15% of MVC) for the muscle current. Increasing current levels caused substantial muscle contractile forces for the muscle current (33% of MVC at the 75% current level, 53% of MVC at the 90% level) and resulted in a significant three-way interaction at both levels.

To further assess the contribution of the pain descriptor factor on the four-way interaction, the coping style x pain descriptor interaction was assessed for each current type at the 90% level of tolerance. This analysis resulted in a significant interaction for the afferent current (F = 10.84; df = 1,28; P = .003), indicating that to the blunters and monitors, pain types were dearly distinguishable. The interaction for the muscle current approached significance (F = 3.6; df = 1,28; P=.068).

Discussion

The four-factor ANOVA revealed a significant four-way interaction (coping style x current type x pain descriptor x current level), indicating that the rate of increase in patient perception of pain associated with increasing levels of NMES is dependent on the preferred coping style of the individual, whether there is an NMES - elicited muscle contraction, and the dimension of pain being assessed. Electrically elicited muscle contractions appear to contribute uniquely to the pain reports by individuals as evidenced by the simple-effects analyses, which revealed significant coping style x current type x pain descriptor interactions only at levels at which substantial muscle contractile forces were elicited.

Researchers studying the discomfort associated with NMES have concluded that there is a great deal of individual variability associated with a subject's responses to different current forms.[7-10] The results of these studies have led researchers to suggest that a variety of current forms be made available so that clinicians can, through trial and error, assess the most efficient current form for each patient.[7] The most efficient current form is the current form that affords the greatest force of muscle contraction with the least amount of discomfort to die patient. This trial-and-error procedure takes on the characteristics of a lottery and results in a very cost-inefficient means of determining the current form that a patient will best tolerate. Therefore, accounting for individual variability becomes the challenge in future work in the area of discomfort and NMES.

With previous research based on varying current forms, investigators focused on the interface between the electrode and the patient's skin in search of the source of patient discomfort.[7-10] Higher current levels resulted in both increased sensation (including pain) and increased muscle contractile forces, and the objective of the previously mentioned research was to find the combination of current characteristics that would minimize pain receptor stimulation but maximize muscle stimulation.

The results of this experiment indicate that NMES-associated discomfort cannot be adequately defined by previous explanations that focus on the barrage of nociceptive afferent input that results from transcutaneous stimulation. Specifically, the results of this experiment indicate that an electrically elicited muscle contraction contributes substantially to the sensation of discomfort, especially at higher current levels. Because higher levels of stimulation have been thought to be most therapeutic in terms of muscle strengthening, this finding has implications for therapists using NMES for muscle-strengthening regimens.

One of the major findings from this experiment is that, in addition to discomfort from the afferent stimulus as a result of transcutaneous current being applied, there was a substantial degree of discomfort related to the electrically elicited muscle contraction. During higher current levels (ie, 75%, 90%), the subject's muscle contracted quite forcefully. For example, at the 90% level, contractile force was greater than 50% of the force that could be generated from an MVC (Fig. 5).

Many of the subjects commented that the muscle contraction was a limiting factor in their tolerance to the stimulation. The results of the simple-effects analyses can be argued to substantiate this claim. As current levels increased, so did the magnitude of the electrically elicited muscle contractions (Fig. 5). In comparing the three-way interactions of coping style x current type x pain descriptor at the three current levels, there was no significant three-way interaction at the 60% current level (Fig. 2). The contractile forces at the 60% current level were minimal for the muscle current, averaging only 10% to 14% of maximal volitional effort for both groups (Fig. 5). Thus, afferent input from both the muscle current and the afferent current electrode placements appeared to be equivalent in terms of pain and unpleasantness ratings between groups. Therefore, it appears that the effects of the two current types were equivalent at the 60% level of current, at which minimal muscle contractile forces were elicited.

The same three-way interaction was significant, however, at both the 75% and 90% current levels (Figs. 3 and 4). The contractile forces elicited by the muscle current at the 75% and 90% tolerance leverage averaged 33% and 53% of MVC, respectively, for both groups (Fig 5). Thus, when the muscle current

levels were high enough to cause concomitant muscle contractions of relatively higher forces, the descriptions of discomfort were distinctively than if equivalent levels of afferent current were perceived but without concurrent high force muscle contraction.

The implications of the findings of muscle contractile forces contributing uniquely to NMES discomfort are substantial. Changing current characteristics eg, waveform, burst rate, frequency) was thought to improve subject comfort by minimizing the afferent component of discomfort (eg, nociceptive input). Clinicians should be aware, however that factors other than the amplitude and the type of the applied current also contribute to NMES-associated discomfort, regardless of the current type. This is especially true when high muscle contractile forces are desirable, as in most muscle-strengthening techniques. A strategy that focuses on changing current characteristics to improve subject comfort will not affect the discomfort that is associated with muscle contractile forces. By recognizing that muscle contractile forces will contribute to the discomfort that subjects feel, clinicians can intervene with different strategies to assist subjects in tolerating NMES procedures. An example of an alternative coping strategy is providing the subject with information concerning sensations that can be expected specific to an electrically elicited muscle contraction.

The implications of the group differences for clinicians using NMES can also be argued to be substantial. Using specific methods of preparation tailored for particular coping styles has been shown to be successful in minimizing adverse responses in patients undergoing potentially aversive medical procedures.[26] In the same light, preferred coping styles of potential subjects should be sought out by clinicians using NMES, especially given the ease of administration of the MBSS. If a clearly preferred coping style is ascertained, then clinicians can impose specific strategies that take advantage of this coping style. For example, if a subject is a monitor, then providing information to the subject about all aspects of die procedure would be of paramount importance. In light of previous findings, information concerning muscle contractile sensations would be especially important to explain to subjects whose preferred coping style is information seeking. Conversely, if a subject is a blunter, providing a means to enhance a blunting style would be important. For example, providing a quiet setting in which distractive techniques are available would be important for any subject whose preferred coping style during aversive procedures is to blunt.

In addition to suggesting that interventions be tailored to a subject's coping style, these results may further suggest that specific interventions be used depending on a subject's description of the intensity and unpleasantness of the stimulus. Gracely and colleagues[19,20] have demonstrated that different pharmacological interventions have distinctive effects on either unpleasantness or intensity of clinical pain. For example, administration of an opioidlike analgesic in one study[20] resulted in less intensity of pain and no effect on unpleasantness, whereas in another study[19] an antianxiety drug resulted in subjects reporting less unpleasantness of pain and no change in intensity. As a general rule, blunters in our study found muscle stimulation current to be equally as intense as it was unpleasant, whereas monitors found muscle current to be predominantly intense. All of the subjects in our study were "healthy" individuals who were not taking any analgesic medications. Many patients who would benefit from NMES are also prescribed various analgesic agents, however, especially those who are seen shortly after certain surgical procedures. Perhaps dispensing specific, short-acting pharmacological agents or timing those agents already being used would improve a subject's tolerance to NMES procedures.

Limitations

This study was conducted with a theoretical question in mind. Consequently, external validity was not judged a priority and subjects with no known pathology were used. We suggest caution in generalizing the results of this study to patients with pathological conditions, particularly those with central or peripheral nervous system disorders. Additionally we wish to reiterate that the results only suggest interventions to improve patient tolerance to electrical stimulation by tailoring a setting to a subject's coping style. Further testing, particularly of patients serving as subjects, is needed to corroborate this notion.

Summary

The results of this investigation suggest that NMES-associated discomfort is dependent on (1) a subject's preferred coping style, (2) the dimension of pain being described, and (3) whether the current level is high enough to cause a forceful muscle contraction. These results indicate that (1) behavioral styles will affect how a subject will characterize the discomfort associated with NMES and (2) involuntary muscle contractile forces contribute uniquely to the discomfort felt from NMES. Further, these results suggest that interventions tailored to a preferred coping style may increase a subject's tolerance level to NMES and thus provide a more beneficial treatment.

(*) Versa-Stim [TM] 380, Electro-Med Health Industries Inc, 6240 NE 4th Ct, Miami, FL 33138. (dagger) Cybex [R] II, Cybex, Div of Lumex Inc, 2100 Smithtown Ave, Ronkonkoma, NY 11779. (double dagger) Versa-Stim, CONMED Corp, 310 Broad St., Utica, NY 13501. (section) SYSTAT Inc, 1800 Sherman Ave, Evanston, IL 60201-3793.

Commentary

Dr Delino and colleagues have presented information that challenges users of neuromuscular electrical stimulation (NMES) to rethink their protocols and their interactions with patients. They also point the way for clinical research that will directly test their observations about NMES in a more clinically relevant context. My discussion, however, will not focus on the issues related to electrical stimulation, but rather on what I see as the more global implications of this study. At one level, Dr Delitto and coauthors have done little more than under score what we already know - that people are different and that their individual differences need to be considered when we plan and implement treatments. The study elucidated a theory and, because of the lack of a patient group, does not yet indicate whether treatment would be affected by individual coping-style differences. The finding, however, that a person can tolerate a different dosage of a commonly used therapeutic modality if the intervention is tailored to his or her preferred coping style, has implications for almost all areas of physical therapy practice.

Many times within the pages of this journal and elsewhere we have seen discussions tons about what makes one clinician better than his or her peers.

References

[1] Delitto A, Robinson AJ, Clinical Electrophysiology: Electrotherapy and Electrophysiologic Testing Baltimore, Md: Williams & Wilkins; 1989:95-138. [2] Delitto A, Snyder-Mackler L. Two theories of muscle strength augmentation using percutaneous electrical stimulation. Phys Ther. 1990,70:158-164. [3] Selkowitz DM Improvement in isometric strength of the quadriceps femoris muscle as a result of electrical stimulation. Phys Ther. 1985;65:186-196. [4] Currier DP, Mann R. Muscular strength development by electrical stimulation in healthy individuals. Phys Ther. 1983;63:915-921. [5] McMiken DF, Todd-Smith M, Thompson C. Strengthening of human quadriceps muscles by cutaneous electrical stimulation. Scand J Rebabil Med. 1983;15:25-28. [6] Laughman RK, Youdas JW, Garrett TR, et al. Strength changes in normal quadriceps femoris muscle as a result of electrical stimulation. Phys Ther. 1983;63-494-499. [7] Delitto A, Rose SJ. Comparative comfort of three waveforms used in electrically eliciting quadriceps femoris muscle contractions. Phys Ther. 1986;66:1704-1707. [8] Reisman MA. A comparison of electrical stimulators in eliciting muscle contractions. Phys Ther. 1984;64:751. Abstract. [9] Grimby Wigerstadt-Lossing I. Comparison of high- and low- frequency muscle stimulators. Arch Phys Med Rehabil. 1989;70-835-838. [10] Baker LL, Bowman BP, McNeal DR. Effects of waveform on comfort during neuromuscular electrical stimulation. Clin Orthop. 1988;233;75-85. [11] Averill JR. Personal control over aversive stimuli and its relationship to stress. Psychol Bull. 1973;80;286-303. [12] Miller SM. Controllability and human stress: method, evidence and theory. Behav Res Ther. 1979;17:287-304. [13] Thompson S. Will it hurt less if I can control it? A complex answer to a simple question. Psychol Bull. 1981;90:89-101. [14] 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. [15] Tan S. cognitive and cognitive-behavioral methods for pain control: a selective review. Pain. 1982;12:201-228. [16] Melzack R, Torgerson WS. On the language of pain. Anesthesiology. 1971;34:50-59. [17] Gracely RH. Pshychophysical assessment of human pain. In: Bonica JJ, ed. Advances in Pain Research and Therapy. New York, NY: Raven Press; 1979:805-824. [18] Gracely RH, McGrath P, Dubner R. Ratio scales of sensory and affective verbal pain descriptors. Pain 1978;5:5-18. [19] Gracely RH, McGrath P, Dubner R. Validity and sensivity of ratio scales of sensory and affective verbal pain descriptors: manipulation of affect by diazepam. Pain. 1978;5:19-29. [20] Gracely RH, McGrath P, Dubner R. Narcotic analgesia: fentanyl reduces the intensity but not the unpleasantness of painful tooth pulp sensations. Science. 1979; 203:1261-1263. [21] Gracely RH, Dubner R, McGrath P, Heft M. New methods of pain measurement and their application to pain control. Int Dent J. 1978; 28:52-65. [22] Lodge M. Magnitude scaling: quantitative measurement of opinion. In: Sullivan JL, ed. Series: Quantitative Applications in the Social Sciences. Beverly Hills, Calif: Sage Publications Inc; 1981-24-41. [23] Stevens SS, Galanter EH. Ratio scales and categorical scales for a dozen perceptual continua. J Exp Psychol. 1957;54:377-411 [24] Rosenthal P, Rosnow RL. Essentials of Behavioral Research: Methods and Data Analysis. New York, NY: McGraw-Hill Inc; 1978:336-338. [25] Wilkinson L. SYSTAT: The System for Statistics. Evanston, III: SYSTAT Inc; 1990. [26] Watkins LO, Weaver I, Odegaard V. Preparation for cardiac catheterization: tailoring the content of instruction to coping style. Heart Lung. 1986;15:382-389.