Changing chronic pain experience

165

Pain, 32 (1988) 165-172 Elsevier PA1 01136

Changing chronic pain experience H.C. Philips Department

of Psychology,

Shaughnessy

Hospital,

Vancouver, BC V6H 3Nl (Canada)

(Received 18 August 1986, revised received 22 June 1987, accepted 9 July 1987)

Training in relaxation is a prominent component in multi-disciplinary approaches to the management of chronic Spain, although its specific potency in modulating pain has not yet been established. Using a within-session design, the specific effects of relaxation induction were evaluated in a group of benign chronic pain patients (n = 24), and compared to similar patients undergoing a control procedure (n = 22). The results indicated that irrespective of the focus of pain complaint, induction of relaxation over a 20 min period led to significant and sizeable reductions in both sensory and affective pain experience. The overall intensity of pain was significantly reduced. These effects were shown to be reliable and independent of onset level of sensory experience or depression. No significant changes occurred in the control group over a comparable period. The implications of these results were discussed with respect to the process by which relaxation reducespain. Key words: Relaxation; Chronic pain; Moderation of pain

Introduction Relaxation is a common treatment component in multi-faceted approaches to the management of chronic pain. These pain management programmes have proved sufficiently potent [14,15,19] to encourage analysis of the contributions made by the treatment components. Theoretical justifications vary as to the reasons for the inclusion of relaxation training (muscular relaxation/reduction of sensory input, control and calming of mental state, etc.), but its importance as a treatment technique is well established [l&24,29,32]. Despite the prominence and overall utility [34] of relaxation in managing chronic pain, its specific power in modulating chronic pain experience has not been established. Patients at a behaviourally oriented pain clinic rate a relaxation strategy

Correspondence

to:

chology, Shaughnessy Canada. 03043959/88/$03.50

H.C. Philips, Department of PsyHospital, Vancouver, BC V6H 3N1,

highest in utility over all other strategies taught in a psychologically directed outpatient service [27]. To what extent this reflects the technique’s capacity to lower episodes of pain experience and/or provides a greater sense of self-control of pain [10,11,27] is not clear. The aim of the current study was to evaluate the specific effects of relaxation induction on the experience of constant, chronic, benign pain in a population of individuals attending a chronic pain clinic. Thus, the within-session effects of this procedure upon pain experience could be assessed. Operant approaches [cf., 12,291 have focused attention upon the assessment and modification of the behavioural indices of chronic pain (activity levels, complaint levels, etc.). Clinics with a pressure to prove themselves cost-efficient have included employment measures and analgesic use, etc. However, paramount to the applicant to the pain clinic is the extent to which he/she will be able to influence, attenuate, or modulate the pain experienced.

0 1988 Elsevier Science Publishers B.V. (Biomedical Division)

166

Fortunately, in the last few years, progress has been made in the measurement of pain experience. Working with the useful preliminary step made by Melzack and Torgerson [22] in the development of the McGill Pain Questionnaire, a new scale has recently been produced which further distinguishes aspects of sensory and affective experience [16,17] and enables the comparison of repeated assessments. Although developed for a headache population, it uses adjectives from the McGill Pain Questionnaire and is suitable for any type of pain. The scale reliably distinguishes 5 separable clusters or aspects of sensory pain experience and two clusters or aspects of affective pain experience. This scale has been found to be reliable and valid 1171. With a scale of this sort, it becomes feasible to examine the specific effects of a procedure such as relaxation on pain experience itself. The feasibility of‘ such an approach was demonstrated in a study of Melzack and Perry [21]. Using the McGill Pain Questionnaire, they demonstrated significant changes in both sensory and affective dimensions of chronic pain as a result of alpha feedback and hypnotic training. Component analyses of psychological treatment approaches have yet to be published, probably because of complex statistical problems raised by the assessment of a multi-dimensional phenomenon treated with a multi-faceted approach. In the meantime, it is feasible to examine individual components of the treatment approaches and assess their strengths in controlling pain experience per se. A start is made in this study by limiting the evaluation to within-session effects of relaxation upon constant chronic pain experience.

TABLE

In all, 46 chronic pain patients attending chronic pain management service were studied. Twenty-four of these patients (experimental group) were evaluated before and after a 20 min relaxation induction, while twenty-two (control group) were evaluated before and after a 20 min period in which they listened to a didactic presentation on the nature of chronic pain. Table I below provides details of the age, sex distribution. pain type. pain chronicity, and average pain intensity in the two groups.

Design and instruments The Headache Scale [16,17] was given to these patients before and after either the induction of relaxation or the didactic presentation, in order to get an evaluation by each patient of his/ her pain experience (occasion 1). This scale consists of 30 adjectives commonly utilized by those in pain, 27 being selected from the McGill Pain Questionnaire [22]. Each adjective can be rated by the subject from 0 to 3 (0 = not at all; 1 = mild; 2 = moderate; 3 = severe). Cluster analysis has revealed 7 distinguishable clusters, 5 of which are sensory and the remaining 2 affective. These clusters have been labelled as follows: sensor) ‘ache’ (Sl), ‘sharp’ (S2). ‘tight’ (S3), ‘autonomic’ (S4). ‘dull’ (S5): affective ‘discomfort’ (Al), ‘anxiety/depression’ (A2). This scale allows a direct quantitative comparison of sensory and affective components when the questionnaire is used

I

AGE, SEX DISTRIBUTION, CHRONICITY, SEVERITY, INTO RELAXATION (n = 24) AND CONTROL GROUPS

Relaxation

group (n = 24)

group (n = 22)

AND CHRONIC (n = 22)

PAIN

TYPE

OF SUBJECTS

(n = 46) DIVIDED

Type

chronicity

Severity

(F/M)

Back

Head

Other

(yr)

(O-5)

40.12 (12.04)

20/4

41%

29%

30%

1.7 (6.80)

2.6 (0.84)

38.0

18/4

31%

31%

38%

10.5 (8.65)

(@.%)

Age (yr)

Control

Subjects

(9.3)

Sex dist.

2.63

167

on repeated occasions. The latter is not possible using the McGill Pain Questionnaire. In addition to this scale, subjects were asked to give overall pain intensity ratings using an analogue scale from 0 to 5 (the PPI-McGill Pain Questionnaire: 0 = no pain to 5 = intense and incapacitating pain). The experimental group (24) cases seen before and after the relaxation induction were assessed in an exactly replicable manner 1 week later using the same induction procedure (occasion 2). These 2 occasions will be referred to as occasion 1 and occasion 2. The control group (pre- and post-didactic presentation) was evaluated only on occasion 1 and provides the control data for this study.

Roeedure

Relaxation induction followed the guidelines of Bernstein and Borkovic [4], although a simplified 9 muscle group method was used on both occasions (transcript available from the author). Emphasis was also placed on slow, regular, diaphragmatic breathing which was to be used throughout. None of the patients had received any training in relaxation prior to this first session. All patients, irrespective of group (control or experimental), were asked to fill in the Headache Scale at the onset in order to indicate their current pain level and its qualities. Immediately post-intervention (relaxation induction or didactic on pain, both of 20 min duration), they were asked again to use the scale to indicate their current pain level and its qualities. Neither group was given any expectations with respect to immediate effects of the interventions, both being presented as important ingredients in a 9 week programme for the management of chronic pain. The Headache Scale was analysed to give scores (range O-3) for each of the 5 sensory clusters and the 2 affective clusters [16].

typical of cases referred for treatment [27] both in the sex distribution, chronicity, and in pain intensity characteristics. The average severity estimates are obtained from 1 week of self-monitoring (O-5 ratings using the pain diaries [6]). They demonstrate the intensity of the daily chronic pain problem endured by these individuals. No significant difference was found between the two groups on any of these measures. (2) Pre-intervention differences

Below can be found a table of means (Table II) for each of the sensory and affective clusters measured by the Pain Questionnaire, and the pain intensity estimates (PPI O-5 scale) for the experimental groups on 2 separate occasions and for the control group. For each occasion, the null hypothesis was tested: that the mean (vector) in the experimental group, pre-intervention, was the same as that of the control. For the sensory and affective measures (considered as multi-variate vector responses), 2-sample Hotelling T2 tests were performed. Occasion 1: sensory vector Hotelling T2 = 5.27, P = 0.45; affective vector Hotelling T2 = 4.60, P = 0.12. Occasion 2: sensory vector Hotelling T2 = 2.87, P = 0.76; affective vector Hotelling T2 = 0.85, P = 0.66). Thus, no evidence was found of any significant differences between the

TABLE

II

PRE-INTERVENTION: MEANS FOR SENSORY @l-5), AFFECTIVE CLUSTERS (Al, 2) AND PPI (pain intensity) ON THE TWO OCCASIONS FOR THE EXPERIMENTAL GROUP AND FOR THE CONTROLS Means at ‘ pre’: Experimental Occasion

1

Control Occasion

2

Results

Sl s2 s3 S4 SS

1.21 0.72 1.26 0.64 0.92

0.93 0.52 1.17 0.50 0.88

0.90 0.46 0.84 0.50 0.82

(I) Experimental and control group comparability

Al A2

1.85 0.86

1.48 0.75

1.32 0.58

The characteristics of the 2 groups of chronic pain cases can be found in Table I. They are

PPI

3.05

2.40

2.43

168

TABLE III MEAN (vectors) OF DIFFERENCE BETWEEN PRE AND POST SCORES FOR THE TWO GROUPS (experiment and controls) ON OCCASION 1

Sensory

Affective PPI

hp.

Control

Sl

-0.400

s2 s3 s4 s5

- 0.356 - 0.571 -9.211 - 0.375

0.064 0.091 - 0.107 0.062 - 0.048

Al A2

- 0.963 - 0.421

0.119 0.058

-0.914

0.245

experimental and control subjects on either oCcasion on the sensory or affective clusters. The &in intensity (PPI) was analysed using a univariate, 2-sample t test (t = 1.479, df= 39, P = 0.09). This test also showed no evidence of any experimental/control differences pre-intervention.

(3) The eflects of relaxation induction To evaluate the experimental effect of relaxation, the experimental and control groups were compared at occasion 1, before, and after the 20 min interventions (relaxation induction versus didactic control). Taking a post-minus-pre vector of differences, the null hypothesis assessed that the mean vector of differences of both groups would be the same, Hotelling T2 (2 samples) were performed for both sensory and affective measures, while a univariate Z-sample t test evaluated PPI changes. All tests were significant, indicating a specific shift in sensory and affective components of pain experience, as well as the overall PPI score (sensory Hotelling T2 = 24.14, P = 0.003; affective Hotelling T2= 26.9, P=O.OOO; PPI t =4.81, df= 29, P = 0.001 (two sided) in the experimental group). The consistently negative differences on all measures for the experimental group (see Table III) indicate that relaxation leads to reduction in pain experience consistently across all measures. The control group shows very little change, most of which is in the direction of increased pain.

A comparison of the experimental group on occasion 2 with the controls allows an evaluation of the repli~ability of the above results. Exactly the same type of analyses were performed 3s previously (see 3 above). Again, relaxation had a significant effect on ail 3 sets of measures (sensory Hotelling T* = 11.36, P = 0.013: affective Hotelling T2 = 43.06. P = 0,000: PPI t = 4.09, Q’=- 39. P = 0.002 (two-sided)), A replication of the experimental effect was evidenced. (5) The difference between occasions in a relaxation effecr

If the magnitude of relaxation effect were the same on the 2 occasions, the difference (post minus pre) on the 2 occasions would be the same. To assess this, the difference between differences on the 2 occasions was taken. If the relaxation effect was the same in magnitude, this last measure would be predicted to be zero. The Hotelling T2 tests for both sensory and affective measures were as follows: sensory Hotelling T* = 22.17, P = 0.017; affective Hotelling T2 = 1.09, P = 0.60. Thus the magnitude of effect for the sensory component differed between occasions while the affective component showed no change. A comparable approach on the pain intensity (PPI) revealed the magnitude of the relaxation effect to be the same on the 2 occasions for this measure (t = 0.45, df = 15. P = 0.66 (two sided)). (6~ Differences between pain groups with respect to the effect of relaxafion No significant differences were found between the 3 types of pain in this population of chronic pain patients (back, headache, and other - facial, pelvic, multiple, etc. (Table I)) in their response to relaxation induction. (7) Aspects of the pain experience The inventory used allows analysis of the various clusters that comprise the pain experience dividing into sensory and affective aspects. Pre-intervention, the affective cluster was significantly higher than the sensory cluster score when occasions 1 and 2 were summed (F = 67.8, 1, 23. P -z 0.0000). The sensory factor total at onset was

169

TABLE IV INTERRELATION OF CHANGE SCORES FOR SENSORY AND AFFECTIVE PAIN EXPERIENCE TOTALS OF PAIN INTENSITY RATINGS Sensory Sensory total Affective total

-

Affective

PPI

r = 0.72

r = 0.56

P < 0.000

P < 0.02

r = 0.39 ns

PPI

not significantly correlated, however, with changed scores (post-pre-relaxation). However, for the affective total, the association was significant (r = -0.52, P < 0.01). Thus the higher the affective scores at onset, the smaller the reduction produced by relaxation induction. The table above (Table IV) gives the correlations between change scores. Sensory and affective component changes have 52% of variance in common. The relationship of overall intensity ratings and sensory/affective component change scores is much weaker. Pain intensity changes appear to be unrelated to the affective shift that occurs. The mean depression score for the total group of experimental subjects was 16.17 on the Beck Depression Inventory thus placing them in the mild depression range [2]. The depression scores were significantly associated with the affective and sensory pain totals prior to occasion 1 (affective total: r = 0.67, P < 0.0000; sensory cluster: r = 0.61, P < 0.001). Higher depression scores were found with higher pain levels, as assessed on the sensory or affective dimension. However, no relationships between depression scores and change scores were found for either dimension. Discussion

The characteristics of the chronic pain patients evaluated in this study are representative of the level and type of distress found in individuals who are battling constant pain experience over extended periods of time. The larger proportion of females in this study reflects the clinic’s exclusionary criteria which screen out all individuals currently involved in litigation.

The inventory used to assess pain experience allows the differentiation of sensory and affective aspects of pain experience. These individuals show a much higher level of affective pain experience than sensory, though both are at a clinically significant level. The raised affective component emphasizes the importance of pain management approaches focusing upon, and developing, techniques to help individuals to deal with this emotional reaction to prolonged pain. As has been noted frequently in the past [cf., 301, these chronic pain patients reveal mild levels of depression. The latter is significantly and positively correlated with the level of pain experience be it affective or sensory in quality. Irrespective of the focus of the pain complaint, the induction of relaxation over a 20 min period in previously untrained chronic pain patients led to a significant and sizeable reduction both in sensory and in affective pain experience. In addition to the lowering of the level of both of these dimensions, the overall intensity of pain (PPI) was also significantly reduced. The use of the inventory allowed this more usual and blunt evaluation of pain to be extended to the many types of experience that are incorporated in an intensity index. It appears that the reductions produced by relaxation occur across all the clusters assessed by the inventory, be they sensory or affective in quality. Gracely et al. [13] have been able to show that the common minor tranquilliser (diazepam) can affect pain experience in a specific way. The sensory component of the pain experience remains unchanged while the affective emotional reaction to pain was significantly reduced. In this case, with a relaxation induction, both the sensory and the affective reaction were significantly affected suggesting that relaxation induction entails more than a minor tranquillising effect. It appears to significantly alter the sensory/ discriminative dimension of chronic pain, influencing as it does many types of sensation: tightness, aching, sharpness, etc. This experimental effect produced by the simple induction of relaxation was shown to be reliable by its replication 1 week later in the experimental group. The size of effect on the affective component and on the pain intensity rating was comparable on these 2 occasions, although there is

170

some suggestion that there is some change in the focus of the sensory change on the second occasion. Although the shift remains significant, the relevant contribution of the 5 sensory components varied to some extent. To what extent this is a learning effect, as the subjects become more proficient at relaxation induction, will need to be further investigated. Chronic pain patients with different pain problems benefitted equally from the induction of relaxation procedures. There is no suggestion that one pain type benefitted more from learning to relax. This fact has implications for those working with chronic pain patients and wishing to utilize a group format. It appears to be a technique which will be useful for all chronic pain patients and, as it can readily be induced in a group, can be utilized as an important strategy which will help individuals in that group irrespective of the differences in pain focus. The size of the response to relaxation was independent of the onset levels of sensory experience or of depression. Thus, extremely high levels of sensory experience or depressive scores, at least in the mild range, would not appear to be appropriate reasons for excluding individuals from pain management approaches that utilize a relaxation technique. On the other hand, the affective change produced by relaxation was smaller when the affective component was higher. Individuals with extremely high affective scores pre treatment may need more than relaxation training to manage their problems. The high correlation between sensory and affective change suggests the integrated way in which relaxation influences pain experience. The failure of pain intensity rating changes to correlate highly with sensory or affective changes underlines the bluntness of this simple rating scale to assess the changes in experience. Interestingly, this overall rating seems to be most influenced by the sensory dimension of pain experience, and its bluntness may well be predominantly related to its inability to accurately reflect the changes in affective pain experience. Prior to the relaxation induction, the overall pain intensity (PPI) was significantly related to affective total (r = 0.59; P < 0.005) and to sensory total (r = 0.65; P c 0.001). Although

reflecting to some extent sensory and ;tffective qualities, the weakness of PPI is much more evident in its inability to reflect changes in the affective emotional reaction to pain. at least in this context. In summary, it appears that relaxation without former training can have a specific muting or attenuating effect upon persisting chronic pain experience. This suggests the importance of including progressive relaxation techniques in the training battery for chronic pain patients, both as a general component in training as well as a specific episodic control strategy which can be utilized when pain levels peak beyond a point where they can be disregarded. Further training in relaxation may well produce more sizeable reductions as well as quicker ones. The relative potency of relaxation at difference levels of pain experience would be an interesting further investigation to clarify the relative potency of relaxation induction for these patients. To further clarify the process by which progressive relaxation leads to changes in pain experience concurrent psychophysiological assessment and ratings of subject arousal/anxiety must be included in future studies. This additional information would help unravel the way in which such a simple intervention as relaxation modulates a long-standing and continuing pain problem. Some researchers have presumed relaxation induction to be effective in its capacity to reduce sensory discriminative input [1,32]. However, as relaxation has been shown to reduce arousal and anxiety. it is possible that relaxation produces its effect via modulation of the affective/motivational dimension of pain experience. Recently it has been shown that pain perception is significantly influenced by the induction of relevant anxieties [35]. The correlation between sensory and affective pain experience would suggest that in order to further clarify this issue, independent measures of anxiety and muscle tension would be a useful step to take in teasing out the focus of effect produced by relaxation induction. In conclusion, this study has made clear the relevance and importance of training any type of chronic pain patient in the use of relaxation procedures. Relaxation produced consistent reductions

171

in all aspects of the pain experience, be they sensory or affective. As this study evaluated the effects of only the first 2 sessions of relaxation training, it is possible that the speed of reduction and/or its size may be further increased as training continues. The demonstration to patients that they can, over a short period, modulate their distress by a simple technique is of great potency at the onset of training.

Acknowledgements

Thanks are due to Tania Deans for preparing the manuscript, and to the Statistics Department at the University of British Columbia for their advice in data analysis.

References of Chronic Headache, 1 Bakal, D.A., The Psychobiology Springer, New York, 1982. M., Mock, J. and 2 Beck, A.T., Ward, C.H., Mendelson, Erbaugh, J., An inventory for measuring depression, Arch. gen. Psychiat., 4 (1961) 561-571. of Subjective Responses, 3 Beecher, H.K., The Measurement Oxford University Press, New York, 1959. 4 Bernstein, D.A. and Borkovic, T.D., Progressive Relaxation Training. A Manual for Helping Professions, Research Press, Illinois, 1973. E.W. and Andrasik, F., Psychological assess5 Blanchard, ment and treatment of headache: recent developments and emerging issues, J. consult. clin. Psychol., 50 (1982) 859-879. T.H., Stoyva, J.M. and Adler, C., Subjective 6 Budzynski, assessment of pain and its relationship to the administration of analgesics in patients with advanced cancer, J. psychosom. Res., 10 (1973) 203-308. 7 Chapman, C.R., Pain: the perception of noxious events. In: R.A. Stembach (Ed.), The Psychology of Pain, Raven Press, New York, 1979, pp. 169-202. pain management program: 8 DeGoode, D.E., A behavioural expanding the psychologists’ role in the medical setting, Prof. Psychol., August (1979) 63-78. 9 Dolce, J., Pain management: a reaffirmation, Behav. Ther., 7 (1984) 38-50. 10 Dolce, J.J., Cracker, M.F., Moletteire, C. and Doleys, D.M., Exercise quotas, anticipatory concern and self-efficacy expectancies in chronic pain, Pain, 24 (1986) 365-372. 11 Dolce, J.J., Doleys, D.M., Raczynski, J.M., Lossie, J., Poole, L. and Smith, M., The role of self-efficacy expectancies in the prediction of pain tolerance, Pain, 27 (1986) 261-272.

Methods for Chronic Pain and 12 Fordyce, W.E., Behavioural Illness, Mosby, St. Louis, MO, 1976. P. and Dubner, R., Validity and 13 Gracely, R.H., McGrath, sensitivity of ratio scales or sensory and affective verbal pain descriptors: manipulation of affect by diazepam, Pain, 5 (1978) 19-29. 14 Guck, T.P., Skultety, F.M., Meilman, P.W. and Dowd, E.T., Multidisciplinary pain center follow-up study: evaluation with a no-treatment control group, Pain, 21 (1985) 295-307. 15 Holroyd, K.A., Recurrent migraine and tension headache. In: K. Holroyd and T. Creer (Eds.), Self-Management of Physical Disease. Developments in Health, Psychology, and Behavioural Medicine, Academic Press, New York, 1985. 16 Hunter, M., The headache scale: a new approach to the assessment of headache pain based on pain descriptions, Pain, 16 (1983) 361-373. M., Hunter, M. and Philips, H.C., The headache 17 Jahanshahi, scale and examination of its reliability and validity, Headache, 26 (1986) 76-82. 18 Lange, P.J., Fear reduction and fear behaviour: problems in treating the construct. In: J.M. S&lien (Ed.), Research in Psychotherapy, American Psychological Association, Washington, DC, 1978. treatment for 19 Linton, S.J., A critical review of behaviour chronic benign pain other than headache, Brit. J. clin. Psychol., 20 (1986) 321-337. R. and Loeser, J.D., Phantom body pain in 20 Melzack, paraplegics: evidence for a central ‘pattern generating mechanism’ for pain, Pain, 4 (1978) 195-210. of pain in the use 21 Melzack, R. and Perry, C., Self-regulation of alpha-feedback and hypnotic training for the control of chronic pain, Exp. Neurol., 46 (1975) 252-269. W.S., The language of pain, 22 Melzack, R. and Torgerson, Anesthesiology, 34 (1971) 50-79. 23 Melzack, R. and Wall, P.D., The Challenge of Pain, Penguin Books, Middlesex, 1982. self-manage24 Mitchell, K.R. and White, R.G., Behavioural ment: an application to the problem of migraine headache, Behav. Ther., 8 (1977) 213-222. B., 25 Newman, RI., Seres, J.L., Yossep, L.P. and Garlington, Multidisciplinary treatment of chronic pain: long-term follow-up of low-back pain, Pain, 4 (1978) 283-292. in Psychother26 Paul, G.L., Insight Versus Desensitization apy, Stanford University Press, Stanford, CA, 1966. treatment on 27 Philips, H.C., The effects of behavioural chronic pain, Behav. Res. Ther., 25 (1987) 365-378. M., The effects of persistent 28 Philips, H.C. and Jahanshahi, pain: the chronic headache sufferer, Pain, 21 (1985) 163-176. L., The behavioral manage29 Roberts, A.H. and Reinhardt, ment of chronic pain: long-term follow-up with comparison groups, Pain, 8 (1980) 151-162. 30 Romano, J.M. and Turner, J.A., Chronic pain and depression: does the evidence support a relationship?, Psychol. Bull., 97 (1985) 18-34. 31 Stembach, R.A., Pain: a Psychophysiological Analysis, Academic Press, New York, 1978.

172

32 Turk, DC., Meichenbaum, D. and Genest, M., Pain and Behavioral Medicine: a Cognitive Behavioral Perspective, Guilford Press, New York, 1983. 33 Turner, J.A., Comparison of group progressive relaxation training and cognitive behavioural group therapy for chronic low back pain, J. consult. clin. Psychol., 50 (1983) 757-765.

34 Turner, J.A. and Chapman, C.R., Psychological mterventions for chronic pain: a critical review. I. Relaxation training and biofeedback, Pain, 12 (1982) 1-21. 35 Weisenberg, M., Aviram, O., Wolf, Y. and Raphael;, N., Relevant and irrelevant anxiety in the reaction tn pain. Pain, 20 (1984) 371-385.