The effects of relaxation with a warning cue on pain tolerance.

Broucek, Mark W. ; Bartholomew, John B. ; Landers, Daniel M. 等

Ryan and Kovacic (1966) demonstrated that contact athletes have a higher level of pain tolerance than do noncontact athletes. Since then there has been a paucity of pain research using athletes. This is surprising given that contact athletes are likely to be exposed to high levels of pain, and much useful knowledge would be gained from their use as subjects. This lack of research is also unfortunate in that, given the demonstrated difference between contact athletes and others in pain tolerance, it is unwise to generalize research using nonathletes to the experience of contact athletes.

The bulk of research that has been done is based on the parallel process theory developed by Leventhal, Brown, Schacham and Engquist (1979). They proposed two competing components of the pain experience; informational, and emotional. The informational component consists of the sensory aspects of the physical, or noxious stimulus. The emotional component consists of feelings of distress or fear, brought about by the uncertainty associated with the noxious stimulus. An individual will experience pain, they suggest, in direct proportion to the amount of attention focused on the emotional rather than on the informational component of the physical reaction to the noxious stimulus (Leventhal et al., 1979).

With parallel process theory as a basis, a number of researchers have attempted to increase pain tolerance through either "redefinition or "distraction" (See McCaul & Malott, 1984, for a review.). Redefinition consists of focusing attention on the informational component of the noxious stimulus, stressing its location and sensory qualities and processing them in a nonemotional manner. It is thought that by so doing, the uncertainty is decreased, thereby reducing the perception of pain. Distraction consists of focusing attention away from both the informational and emotional components of the noxious stimulus. If successful, this will partially exclude the noxious stimulus from processing, and the perception of pain will decrease. For instance, an intervention strategy, such as visualization, may be used so as to decrease the ability of the individual to focus attention on the noxious stimulus.

This model assumes that reacting to a pain stimulus involves controlled rather than automatic information processing (Schneider & Shiffrin, 1977; Shiffrin & Schneider, 1977). In controlled processing, attentional capacity is limited. Thus, focusing attention on the informational component of a pain stimulus (redefinition) would reduce the amount of distress experienced. Similarly, attention focused on an entirely different stimulus (distraction) would reduce attention to both the informational and emotional components of a pain stimulus, also reducing experienced distress.

Both of these approaches are focused on the reaction of the nervous system to a noxious stimulus; either attempting to redefine it, or to ignore it. A person might be better served, though, by attempting to intervene one step earlier and modify the actual amplitude of the physiological response to the stimulus. One such response is an increase in physiological arousal following a noxious stimulus. This increase has been described as resulting from the novelty of the physical stimulus, and the uncertainty about the meaning of the stimulus (Chapman, 1986). It is this type of response which is often processed emotionally, leading to a greater perception of pain. Thus, an intervention that would lower arousal should also decrease the perception of pain.

This kind of intervention differs from distraction. Distraction is an attempt to ignore the afferent reaction to a noxious stimulus. Instead, when the level of arousal is reduced, one aspect of the physical response to the stimulus is diminished. There is, therefore, a weaker afferent reaction to interpret emotionally, and the perception of pain is decreased. Reeves and Shapiro (1983) have shown that subjects trained to lower their heart rates reported lower pain ratings on a cold pressor task. Also, Kaplan, Metzeger and Jablecki (1983) have demonstrated that individuals trained in deep breathing relaxation reported lower pain ratings on an electromyography exam, in which electric shocks are used to stimulate nerves and muscles. Progressive muscle relaxation (Jacobson, 1938), which has been used extensively in helping athletes control their level of arousal in stressful situations, may be another, perhaps more effective, intervention.

McCaul and Malott (1984) have argued that as the intensity of the pain stimulus increases it demands more of one's attentional capacity. Therefore they suggest using distraction with mild pain and redefinition with more intense pain. High levels of pain intensity could also compete for attention with the relaxation task. A high level of pain would then render relaxation ineffective as an intervention. However, if relaxation was implemented prior to the onset of the pain stimulus, before there is competition for attention, it may allow the subject to maintain a lower level of arousal and to experience a less intense reaction to the pain.

Linton and Gotestam (1983) conducted a study which examined this hypothesis. In their investigation subjects were trained in relaxation through the use of three 8-minute tapes. Two groups of subjects were then given a 1-minute signal prior to the onset of the pain stimulus, a cold pressor task, and were instructed to either relax at the signal, or after the stimulus had begun. A control group was given no signal. Results indicated no difference in pain thresholds between groups. However, there are three methodological issues that may invalidate this investigation.

Foremost is the method of training employed and the lack of verification that the relaxation response was both adequately learned and appropriately employed during the study. Linton and Gotestam (1983) used taped instructions to teach the relaxation skill, and merely employed a questionnaire to ensure that the relaxation response was effectively employed. Borkovic and Sides (1979), in their review of the relaxation literature, stress the importance of using live rather than taped training sessions for a sufficient relaxation effect to occur. In the study described in this report, 3, 1 hr progressive muscle relaxation training sessions were conducted by the experimenter. A physiological measure, heart rate, was also used to assess the relaxation response. A further measure, electromyography (EMG) of the frontalis muscle, was used during training to determine that the relaxation response had been sufficiently learned.

Secondly, Linton and Gotestam (1983) allowed only 1 minute for the subjects to relax prior to the onset of the pain stimulus. This time constraint may have been severe enough to place undue pressure on the subjects to relax. Such pressure would not have been conducive to relaxation and could have hindered such a response. In the present study, subjects were allowed 2 minutes within which to relax prior to the pain stimulus.

The final problem concerns the device used to elicit a pain response, the cold pressor task. There is an unavoidable ceiling effect associated with this task. An increase in pain is only experienced up to the point at which numbness occurs. Also, it is not at all uncommon in studies using the cold pressor task for a number of subjects to reach the ceiling level on the tolerance measurement. In addition, being exposed to ice is a routine experience for athletes. Therefore, they may be habituated to the stimulus in a cold pressor task. A gross pressure device, similar to the one employed by Ryan and Kovacic (1966), has been shown to be a reliable method of inducing pain without the limiting factors mentioned above (Brewer, Karoly, Linder & Landers, 1990). The gross pressure device was therefore selected as the method of pain induction.

The purpose of the present study, then, was to examine the effect of relaxation, taught using the progressive muscle relaxation technique, on the pain tolerance of athletes. In addition, the effect of providing a warning signal prior to the onset of the pain stimulus was investigated. It was predicted that those subjects trained in progressive relaxation and provided with a warning cue 2 minutes prior to the onset of the pain stimulus would report a greater pain tolerance than identically trained subjects without a warning cue.

Method

Design

A five condition, nonequivalent control group, repeated measures design was employed (Campbell & Stanley, 1963). Two conditions were composed of contact athletes, trained in progressive muscle relaxation. Athlete subjects were randomly assigned to either relax at the warning signal, or to relax at the onset of the pain stimulus. Two placebo control conditions and a no intervention control condition were composed of students enrolled in physical education classes. Control subjects listened to soothing music instead of receiving relaxation training, and were randomly assigned to groups instructed either to relax at the warning signal, or to relax at the onset of the pain stimulus. The no intervention control group, received no training or music and no relaxation instruction prior to the onset of the pain stimulus.

The ideal design for this experiment would have been a completely randomized 2 (warning signal, no warning signal) X 3 (relaxation training, placebo control, nonintervention) factorial design, employing only contact athletes as subjects. The present design was made necessary by the unavailability of a sufficiently large sample of contact athletes to allow random assignment of athletes to all conditions. In this design, the critical comparison of athletes' pain tolerance with and without a warning signal, after relaxation training, is not confounded with the athlete-nonathlete variable. The effect of relaxation training compared to a placebo control is confounded with the athlete-nonathlete variable, and any main effect of relaxation training on pain tolerance could not be confidently interpreted. However, the nonathlete placebo conditions do allow an assessment of the impact of the warning period on pain tolerance, in the absence of relaxation training. Although subject to some limitations, the present design will allow a confident interpretation of the impact of a warning signal on the effectiveness of relaxation training as an intervention for coping with pain.

Subjects

Subjects were 56 students enrolled in physical education classes at Arizona State University. Contact athletes in the experimental group, who were actively competing on the college level, were recruited from EPE 348, Psychological Skills for Optimum Performance. The control subjects were recruited from both EPE 348 and EPE 346, Psychology of Coaching. All subjects signed an informed consent form approved by the Human Subjects Institutional Review Board.

Pain Device

The gross pressure device (Ryan & Kovacic, 1966) consisted of a flat, rubber coated, football cleat secured to a plastic, foam padded, soccer shin guard. The cleat was placed midway between the ankle and the knee, on the medial surface of the tibia. The sleeve of a standard clinical sphygmomanometer was used to secure the device and induce pressure.

Procedure

Upon entering the laboratory, subjects were informed about the pain test and the general purpose of the study. The subjects then completed a consent form, as well as a questionnaire stating whether or not the subject's heart rate was artificially raised or lowered by drugs or caffeine. The date and time of the session were recorded to allow the posttest to be scheduled as close as possible to the time of day of the pretest. A heart rate monitor (Heart Speedometer, an earlobe sensor manufactured by Computer Instruments Inc.) was attached to the subject, the pressure cleat was secured to the leg, and audio-taped instructions, explaining that they were to attempt to endure as much pain as possible, were played. Subjects were asked to count backwards from 500 by 7s, so as to prevent relaxation attempts before instructed to do so. The trial was then begun.

The sleeve of the sphygmomanometer was inflated at a slow, constant rate until the subject asked the experimenter to remove the pain stimulus. The level of mercury (mmHG) was recorded as the measure of pain tolerance, and the pressure released. To ensure that maximal effort was given on the second trial, the experimenter commented that the resulting score was quite a bit lower than the average group tested. The cleat was then lowered one inch and a second trial was administered.

Heart rate was recorded at the following intervals: (I) every 30 seconds for the 2 min alter the monitor was attached and before the taped instructions; (2) immediately after attaching the pain device; (3) every 30 seconds prior to the pain stimulus; (4) at the beginning of the pain test; (5) at a reading of 100 mm/Hg on the sphygmomanometer (a painful but submaximal level); (6) at maximal pain tolerance; and (7) immediately after the pain device was removed.

After the pretest was completed the athlete subjects were given at least three I-hour progressive muscle relaxation training sessions. Frontalis EMG was recorded (Thought Technologies, EMG 100T), and a criterion EMG level reduction to 3.0 ||micro~volt~ average was established as a minimum requirement for having learned the relaxation response. This level was established by Stovya (1983) as an acceptable level of relaxation. Sessions were administered by the experimenter who was trained in progressive muscle relaxation (Jacobson, 1938). The placebo control groups listened to soothing music tapes which corresponded to the length of the training sessions. The no intervention control group received no training.

After completion of the training the posttest was administered. This was identical to the pretest except for the instructions to relax. The athlete and nonathlete relax at signal groups were instructed to relax at the warning signal, which preceded the pain stimulus by 120 s. The athlete and nonathlete relax at pain groups were instructed to relax at the onset of pain, and no warning signal was given. The posttest procedure for the no intervention condition was identical to the pretest procedure.

Results

Six subjects were eliminated because they reached the highest permitted intensity on the pretest. This resulted in a sample of 10 subjects per cell, equally divided between males and females. EMG of the frontalis muscle was measured to assure that the relaxation response was adequately learned. All athlete subjects reached the criterion of reduction to 3.0 ||micro~volt~ with a mean pre/post effect size of 1.38 (all effect sizes were computed as standardized mean differences). Therefore only the minimum three training sessions were employed.

Heart rate was taken to determine if subjects were relaxing at the appropriate point. The dependent variable was the difference between the heart rate change in the pre-pain period during the pretest session and heart rate change in the pre-pain period during the posttest session. The change score was measured as the difference between heart rate taken 2 min before the instruction tape (baseline), and the average of the four measures taken at 30 s intervals before the initiation of the pain stimulus (pre-pain heart rate). In the posttest, the first of the pre-pain measures was taken when the warning cue was given to the "relax at signal" groups. Table 1 shows that only the athlete, relax at signal group evidenced a heart rate deceleration in the posttest pre-pain period. A 5 x 2 (Group x Pre/Post) ANOVA with repeated measures was conducted on this dependent variable, revealing only a significant groups by pre/post interaction, F(4, 45) = 11.31, p |is less than~ .001, ES = 2.15. A Tukey post hoc test on these differences revealed that the athletes, trained in progressive relaxation, when asked to relax at the signal, showed a significantly decreased heart rate in the pre-pain period of the posttest session, F(4, 45) = 11.31, p |is less than~ .0001.

The pain tolerance level was measured twice at both pre and posttest. The higher of the two measurements was used as the subjects' pain tolerance score. Table 2 shows that only the athletes, relax at signal group evidenced a substantial increase in pain tolerance on the posttest. A 5 x 2 (Group x Pre/post) ANOVA with repeated measures was conducted on these measures, revealing a significant interaction, F(4, 45) = 26.49, p |is less than~ .001, ES = 1.44. A Tukey pest hoc test revealed that the pain tolerance for the athletes, trained in progressive relaxation, and asked to relax at the signal, showed a significantly greater increase from pretest to posttest than all other groups on this measure, F(4, 45) = 26.49, p |is less than~ .0001.

Table 1

Group Means and Standard Deviations for Heart Rate Differences
on the Pre and Posttests

 Pre Post
Groups M SD M SD ES

Nonathlete Control
No intervention 2.73 3.74 .68 3.22 .55
Relax at pain 2.13 4.13 4.10 4.68 -.48
Relax at signal 1.76 2.91 2.70 1.86 -.32

Athlete Experimental
Relax at pain 1.48 4.10 2.35 2.48 -.21
Relax at signal 4.50 4.63 -5.45 1.10 2.15
Table 2

Group Means and Standard Deviations for the Pain Measures on
the Pre and Posttests

 Pre Post
Groups M SD M SD ES

Nonathlete Control
No intervention 210.7 54.46 200.0 48.07 .20
Relax at pain 212.6 54.75 201.0 57.87 .21
Relax at signal 205.5 55.85 189.5 58.28 .29

Athlete Experimental
Relax at pain 234.5 57.61 242.0 56.58 -.13
Relax at signal 190.2 40.63 248.5 31.89 1.44

Discussion

The focus of this study was to examine the effects of relaxation, with and without a warning cue, on the pain tolerance of athletes. It was predicted that the presence of a warning cue would facilitate the use of the relaxation skill in coping with pain, yielding an increase in pain tolerance for the experimental, relax at signal group. The results supported this prediction. Athletes trained in progressive relaxation and provided a 2-minute warning period during which to relax prior to the onset of the pain stimulus had a significantly greater increase in pain tolerance on the posttest than did either identically trained athletes, who attempted to relax at the onset of pain, or untrained students. These results indicate that a warning cue may be necessary to allow relaxation training to serve as an effective intervention for increasing pain tolerance.

The results from the control conditions allow this interpretation to be made more confidently. Comparing the nonathlete placebo control groups to one another, it is apparent that when the relaxation skill is unavailable, a warning signal followed by a 2 minute waiting period prior to the onset of the pain stimulus has no effect on pain tolerance. It is remotely possible that the different patterns of results are due to a difference between athletes and nonathletes in the ability to use the warning signal effectively, regardless of the relaxation skill. An effect of this sort has not been previously suggested by theory or research, and could be invoked here only as a post hoc alternative explanation. It is unfortunate that a sufficiently large sample of contact athletes was not available to be used as subjects in the placebo control conditions. However, in the absence of a compelling alternative account for the pattern of results, the nonequivalent control group design does allow us to infer support for the assertion that a warning cue is a necessary condition for the effectiveness of progressive relaxation as an intervention to increase pain tolerance.

Implementing the relaxation skill may require attention. When a warning cue is available, attention is directed at the task of relaxation and it can be effectively accomplished. Thus, at the onset of the pain stimulus the coping mechanism is in operation. However, without a warning cue, the pain stimulus and the relaxation task compete for attention (McCaul & Malott, 1984) and the subject may not be able to implement an effective level of relaxation. This analysis suggests thai overlearning the relaxation skill may allow it to become effective in the absence of a warning cue, as an overlearned response requires a smaller proportion of the subject's attentional resources.

It is also possible that the relaxation response requires time to develop to a level at which it is effective for coping with pain. By using relaxation to counter the normal response to a noxious stimulus, an increase in physical arousal, relaxation would be considered an opponent process (Solomon, 1980). As such, it may lag behind the process of attending to and experiencing the distress of pain. Thus, even in the absence of a warning cue, trained subjects may show lower pain ratings in response to a pain stimulus of longer duration at a level below the tolerance threshold. In a recent clinical study, relaxation and imagery have been shown to decrease the perception of submaximal pain in osteoarthritic patients (Keefe et al., 1990).

If relaxation can be shown to influence an athlete's perception of pain during task performance, it may prove beneficial in mitigating the negative effects of pain on performance (Brewer, Van Raalte & Linder, 1990). This however has yet to be shown in either athletic or clinical settings. Although Keefe et al. (1990) have shown that relaxation and imagery can reduce the perception of pain in osteoarthritic subjects, it was not shown to effect task performance, measured in their report as speed and ease of daily movements.

This failure to generalize to performance tasks provides an opportunity for future research, especially in athletic settings. Participation in contact and endurance sports carries with it the experience of submaximal pain. Thus, the athlete's perception of pain may be a mere fruitful area of research than pain tolerance. As such, the relationship between relaxation and athletes' perception of pain at submaximal levels should be examined. Similarly, contact sports rarely provide a warning preceding the pain stimulus. This may be a limiting factor for the use of relaxation as an intervention. Therefore, the possible benefit of overlearning the relaxation skill must be examined. Since overlearned tasks require less attention, overlearning relaxation may allow it to be effectively employed without a warning.

Pain perception and overlearning the relaxation skill are two areas into which future pain research can be directed to examine these effects in athletic performance situations. Because one of the most important goals of any intervention is to improve performance, this is a direction which pain researchers must explore. It must be remembered, however, that it is not always appropriate to attempt an intervention to maintain performance in painful situations. A noxious stimulus can provide information about possible tissue damage and the likelihood of incurring a debilitating injury. Thus, while many injuries are minor in nature, the desire to maintain performance must be tempered with the realization that further damage is a possibility. If used prudently, however, relaxation training may enable athletes to reduce both the physical and psychological effects of the pain they encounter.

References

Borkovic, T. D., & Sides, J. K. (1979). Procedural variables related to the physiological effects of progressive relaxation: A critical review. Behavioral Research and Therapy, 17, 119-125.

Brewer, B. W., Karoly, P., Linder, D. E., & Landers, D. M. (1990). Validation of a procedure to induce pain of varied intensities. Research Quarterly for Exercise and Sport, 61, 268-271.

Brewer, B. W., Van Raalte, J. L., & Linder, D. E. (1990). Effects of pain on motor performance. Journal of Sport and Exercise Psychology, 12, 353-365.

Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi-experimental designs for research. Chicago: Rand McNally.

Chapman, C. R., (1986). Pain, perception and illusion. In R. A. Sternbach (Ed.), The psychology of pain, (2nd ed.) (pp. 153-179). New York: Raven.

Jacobson, E. (1938). Progressive relaxation. Chicago: University of Chicago.

Kaplan, R. M., Metzeger, G., & Jablecki, C. (1983). Brief cognitive and relaxation training increases tolerance for painful clinical electromyographical examination. Psychosomatic Medicine, 45, 155-161.

Keefe, F. J., Caldwell, D. S., Williams, D. A., Gil, K. A., Mitchell, D., Robertson, C., Martinez, S., Nunley, J., Beckham, J. C., Crisson, J. E., & Helms, M. (1990). Pain coping skills training in the management of osteoarthritic knee pain: A comparative study. Behavior Therapy, 21, 49-62.

Leventhal, H., Brown, D., Schacham, S., & Enquist, G. (1979). Effects of preparatory information about sensations, threat of pain, and attention on cold pressor distress. Journal of Personality and Social Psychology, 37, 688-714.

Linton, S. J., & Gotestam, K. G., (1983). Applied relaxation (coping) in control of laboratory pain: Effects of signalled pain and instructions of when to relax. Psychological Reports, 53, 467-476.

McCaul, K. D., & Malott, J. M. (1984). Distraction and coping with pain. Psychological Bulletin, 95, 516-533.

Reeves, J. L., & Shapiro, D. (1983). Heart-rate reactivity to cold pressor stress following biofeedback training. Biofeedback and Self-Regulation, 22, 383-390.

Ryan, E. D., & Kovacic, C. R. (1966). Pain tolerance and athletic participation. Perceptual and Motor Skills, 22, 383-390.

Schneider, W. & Shiffrin, R. M. (1977). Controlled and automatic human information processing: I. Detection, search, and attention. Psychological Review, 84, 1-63.

Shiffrin, R. M. & Schneider, W. (1977). Controlled and automatic human information processing: II. Perceptual learning, automatic attending, and a general theory. Psychological Review, 84, 127-190.

Solomon, R. L. (1980). The opponent process theory of acquired motivation. American Psychologist, 35, 691-712.

Stoyva, J. M. (1983). Guidelines in cultivating general relaxation: Biofeedback and autogenic training combined. In J. V. Basamajian, M. D. (Ed.). Biofeedback, principles and practice for clinicians (pp. 149-168). Baltimore: Williams & Wilkins.