- Email: [email protected]
Strategies for coping with chronic low back pain: Relationship to pain and disability
Pain, 24 (1986) 355-364 Elsevier
355
PAI 00844
Strategies for Coping with Chronic Low Back Pain: Relationship to Pain and Disability Judith A. Turner *,** and Stephen Clancy * * Department of Psychiatry and Behavroral Sciences, RP-IO, and ** Department of Rehabilitation Medicine, Multidisciplinary Pain Center, University of Washington School of Medicine, Seattle, WA 98195 (U.S.A.) (Received
22 May 1985, revised received 15 July 1985, accepted
24 July 1985)
Summary Seventy-four chronic low back pain patients in a study assessing the effectiveness of group outpatient cognitive-behavioral and operant behavioral treatment completed the Coping Strategy Questionnaire (CSQ) and measures of pain, depression, and functional disability pre- and post-treatment. The previously reported factor structure of the CSQ was generally replicated, and significant associations were found between use of ignoring and reinterpretation strategies and downtime, between use of attention diversion strategies and pain intensity, and between tendency to catastrophize and physical and psychosocial impairment. Both treatments resulted in significant changes in types of coping strategies used to deal with pain. Increased use of praying and hoping strategies was significantly related to decreases in pain intensity. Decreased catastrophizing was also significantly related to decreases in pain intensity, as well as to decreases in physical and psychosocial impairment.
Introduction There is considerable evidence that the coping responses of individuals to stressors play an important role in adjustment to the stress [5,17]. The question of how people cope with pain, and especially, of what constitutes effective and optimal coping, is of interest to a large number of researchers and clinicians. Coping, in this context, refers to thoughts and behaviors people use to manage their pain or their
Correspondence
0304-3959/86/$03.50
and requests
for reprints
should be addressed
0 1986 Elsevier Science Publishers
to Dr. Turner.
B.V. (Biomedical
Division)
emotional reactions to the pain, so as to reduce emotional distress. It appears that patients with pain use a wide variety of such strategies [Xl, and a number of studies have demonstrated the impact of training in specific coping skills on pain, distress. and other aspects of pain problems. In experimental pain studies, training in the use of cognitive strategies for pain control has been shown to increase pain tolerance and threshold [3] and decrease pain report [7]. Studies in clinical settings have suggested the effectiveness of coping skills training in reducing pain and distress associated with medical and surgical procedures [12,18.23]. Further. coping skills training has been demonstrated in several controlled studies to decrease headache activity [1.13.15] and pain ratings [9.16,22] of chronic pain patients. Although wide variation in degree of physical and psychological disability among individuals with chronic pain has been noted [6], very little is known about how the ways in which people cope with chronic pain may relate to such impairment. Further, although many chronic pain treatment programs emphasize teaching cognitive and behavioral skills to increase patients’ ability to cope with and control pain, empirical data concerning the effects of such treatments on the use of coping skills, as well as concerning the effects of changes in coping strategy utilization on pain and disability, are lacking. In a first step towards examining these issues, Rosenstiel and Keefe [19] described a measure of the extent to which chronic low back pain patients use 1 behavioral and 6 cognitive coping strategies. Three factors of related coping strategies were identified, accounting for most of the variance in responses on this measure, the Coping Strategy Questionnaire (CSQ). Each of the 3 factors was found to be associated with measures of behavioral and emotional adjustment to pain. The present study sought to determine whether this same factor structure would be replicated in a different patient sample, thus indicating that types of coping styles (functionally related coping strategies) as measured by the CSQ could be reliably identified across patients with different demographic and pain characteristics. The present study had several additional objectives: (a) to determine whether Rosenstiel and Keefe’s [19] findings related to the association between CSQ factors and physical and psychological functioning would be replicated in a new sample, (b) to examine whether cognitive-behavioral and operant behavioral treatments, as compared with a waiting list control condition. produced changes in coping strategy utilization as assessed by the CSQ, and (c) to examine whether CSQ changes were associated with changes in pain and disability.
Methods Subjects Subjects were 74 chronic low back pain patients (47 males and 27 females) who completed 1 of 3 conditions in a treatment outcome study. Subjects were referred by physicians or self-referred following media publicity. Criteria for study entry included the following subject characteristics: persistent low back pain for at least 6 months. age between 20 and 65, and married or co-habiting. Exclusion criteria
351
included back surgery within the past year and history of or current cancer or cardiac disease. The mean age of subjects was 46.3 years (S.D. = 10.0). The mean duration of time since onset of first back pain problem was 14.5 years (S.D. = 11.6) and mean duration of current back pain problem was 6.8 years (S.D. = 8.4). Five subjects were receiving financial compensation or anticipated future legal action related to pain. Ten subjects had undergone surgery for back pain. Measures
The Coping Strategy Questionnaire (CSQ) [19] lists 42 strategies for coping with pain, including 6 different types of cognitive strategies (diverting attention, reinterpreting pain sensations, coping self-statements, ignoring pain sensations, praying or hoping, and catastrophizing) and a behavioral strategy (increasing activity level). On the CSQ, subjects rate on a scale from 0 (never) to 6 (always) how often they use each strategy when they experience pain. They also rate how much control they believe they have over pain, and how much they are able to decrease pain, on a scale from 0 (no control/cannot decrease it at all) to 6 (complete control/can decrease it completely). Subjects also completed a Pain Diary [lo] for 1 week, making hourly ratings of pain intensity on a O-10 scale, and recording type of activity and position (sitting, standing or walking, or reclining). For the purpose of data analyses, following the protocol of Rosenstiel and Keefe [19], average pain was defined as the average of 3 subject ratings: (1) verbally reported pain intensity at the time of assessment, rated on a O-10 scale with anchors of ‘no pain’ and ‘pain as bad as it can be,’ (2) the most severe pain in the past week, as recorded on pain diaries, and (3) the least severe pain in the past week, as recorded on pain diaries. Downtime was defined as the amount of time spent lying down or in a reclining position between the hours of 7 a.m. and 11 p.m., as recorded on the pain diaries. The Sickness Impact Profile (SIP) [4] was used as a measure of pain-related physical and psychosocial disability. This behavioral checklist consists of 136 items in 12 areas of health-related dysfunction and yields a total impairment index as well as summary scores of physical and psychosocial dysfunction. The Beck Depression Inventory (BDI) [2], a widely used 21-item self-report measure, was used to assess current severity of depression. Procedure
Subjects completed the CSQ, SIP, and BDI after telephone screening and before random assignment to 1 of 3 conditions: waiting list control, cognitive-behavioral therapy, and operant behavioral therapy. The Pain Diaries were completed over a 1 week period following completion of the other measures, and prior to beginning treatment (in the case of those assigned to a treatment condition). All measures were completed again following treatment or the end of the 8 week waiting list period. Both treatments were conducted in a group format, in 2 h sessions weekly for 8 weeks. Each group consisted of 5-10 subjects, and spouses attended half the sessions in the operant behavioral treatment. Five groups were conducted in each treatment
TABLt
1
MEANS AND STANDARD DEVIArIONS STRATEGY QUESTIONNAIRE Subscales
OF
RAI’INGS
ON
SUBSCALES
Mean
S.D.
1.74 u.94 3.64 2.32 1.94 1.24
1.25 0.94 1.Oh 0.98 1.15 1.11
Behurwrul cqq struregl 1. Increasing activity level
2.40
1.1X
Ef/emwness rutingv 1. Control over pain 2. Ability to decrease
3.30 3.07
1.19 1.14
Co~nrtrw qprng 1. 2. 3. 4. 5. 6.
OF THE
COPING
.rtrrr1qrav
Diverting attention Reinterpreting pain sensations Coptng self-statements Ignoring pain sensations Praying or hoping Catastrophizing
pain
condition, and each group was led by 1 of 5 experienced Ph.D. level clinical psychologists. With one exception, each therapist led both a cognitive and an operant behavioral group. The cognitive-behavioral treatment included training in systematic progressive muscle relaxation, imagery, covert assertion, and the identification and modification of distorted, maladaptive thoughts related to pain and stressful events. The operant behavioral treatment included a regular aerobic walking program, training subjects and spouses in the identification of pain behaviors, instructing subjects to decrease pain behaviors, instructing spouses to not reinforce pain behaviors and to reinforce well behaviors, and communication training.
Results Twenty-one subjects completed the waiting list, 24 the cognitive-behavioral, and 29 the operant behavioral treatment. As no significant differences were found on demographic and dependent measure variables. the group cohorts within conditions were collapsed for purposes of data analysis. Replicution of CSQ factor structure
Differences were observed in the frequency with which subjects used the different coping strategies, as shown in Table I. As was found by Rosenstiel and Keefe [19], subjects reported they rarely reinterpreted pain sensations as a coping strategy (mean = 0.94). Coping self-statements were most frequently reported (mean = 3.64), followed by ignoring pain sensations (mean = 2.32) and praying and hoping (mean = 1.94). Although the overall ability to control and decrease pain was rated as somewhat higher by these subjects than by Rosenstiel and Keefe’s [19] sample, the scores nonetheless were rather low (mean = 3.30 and 3.07 on a scale of O-6).
359 TABLE
II
PRINCIPAL Factor
COMPONENT Component
ANALYSIS
OF THE COPING
measures
Factor 1
Denial of pain Ignoring pain sensations Reinterpreting pain sensations Ability to decrease pain Diverting attention and praying Diverting attention Praying or hoping Increasing activity Helplessness Catastrophizing Ability to control pain Coping self-statements
STRATEGY
QUESTIONNAIRE
loadings 2
3
0.65 0.81 0.74
0.13 0.16 -0.17
- 0.04 0.43 - 0.25
0.16 - 0.21 0.32
0.81 0.75 0.67
-0.10 0.13 -0.12
0.11 0.46 0.05
0.13 -0.11 0.38
0.75 - 0.60 - 0.69
To determine how the different coping strategies were related to one another and to ratings of ability to control and decrease pain, a principal components analysis, using oblique rotation, was performed. Of the factors emerging from this analysis, only those having eigenvalues of 1 or greater were considered. A strategy was included in a factor if (1) it correlated with the factor at a level greater than 0.50, and (2) it had its highest loading on that factor. This analysis produced 3 factors, as shown in Table II, that accounted for 65% of the variance in CSQ responses. This solution was similar to that in Rosenstiel and Keefe’s [19] study, which also found a 3-factor solution, accounting for 68% of the variance. The first factor, accounting for 33% of the variance, was very similar to the first factor reported by Rosenstiel and Keefe [19], which they labeled Cognitive Coping and Suppression. Individuals high on this factor reported that they attempted to suppress pain through ignoring and reinterpreting pain sensations. They also rated themselves as able to some extent to decrease pain, a variable with a very low loading on the Rosenstiel and Keefe [19] component analysis. Unlike the results found by Rosenstiel and Keefe [19], coping self-statements did not load highly on this factor, which we labeled ‘denial of pain.’ The second factor, accounting for 20% of the variance, was similar to the third factor reported by Rosenstiel and Keefe [19]. Individuals high on this factor reported diverting attention, praying and hoping, and increasing activity as methods used to reduce pain. As did Rosenstiel and Keefe [19], we labeled this factor ‘diverting attention and praying.’ The third factor, similar to the factor labeled Helplessness by Rosenstiel and Keefe [19], accounted for 12% of the variance. Individuals high on this factor, which we also call ‘helplessness,’ frequently endorsed items on the catastrophizing subscale and rated their effectiveness in controlling pain as low. In addition, subjects high on this factor rarely used coping self-statements.
TABLE
III
REGRESSION ANALYSES: LOGICAL AND PHYSICAL Measure
of adjustment
I. Average pain
Denial of pain Diverting attention Helplessness 2. Depresston Denial of pain Diverting attention Helplessness
RELATIONSHIP DISABILITY Total R
0.33
and praying
OF COPING
F ratio for R change
F(3.68)
=
2.10
F(l,68)
=
0.17
F(1.
68) =
F(1.68) 0.54
and praying
F(3.
0.53
4. Physical impairment
0.40
=
68) =
TO PAIN AND
Proportion acct. for by 3 factors (S)
6.91 *** 0.36
F(1.68)
=
0.00
F(3,68)
=
6.44 **
=
1.28
F(3.68)
= 3.22 **
R
29 - 0.04 0.05 0.53
68) = 27.02 ***
F(1.68) F(1,68) F(l.68)
Simple
0.04 0.33 - 0.01
0.10
=
PSYCHO-
11
8.28 **
F(l,68) F(1.
3. Dtsability (SIP total) Denial of pain Diverting attention and praying Helplessness
FACTORS
27 0.07 0.23 0.47
= 2.38 = 20.71 *** 16
(SIP physical)
Denial of pain Diverting attention Helplesness
5. Psychosoctal impairment
0.12 0.21 0.32
F(1, 68) = 1.56 F(1,68)= 1.79 F(1, 68) = 8.84 **
and praying
0.52
F(3, 68) = 6.20 ***
27
(SIPpsychosocml)
Denial of pain Diverting attention Helplessness 6. Downtrme Denial of pain Diverting attention Helplessness
0.41 and praying
0.07 0.18 0.48
F(1, 68) = 1.72 F(l, 68) = 1.12 F(1, 68) = 21.78 ***
and praying
F(3,68) F(1, 68) F(1, 68) F(1.68)
= = = =
3.35 ** 6.14 * 2.14 0.75
16 0.31 0.27 0.06
* P < 0.02. ** PC
0.01. *** P < 0.001.
Relationship between coping strategies and physical and psychological functioning We next performed hierarchical regression analyses to examine the relationship between coping strategies and adjustment to pain, after accounting for the effect of duration of pain. The results are shown in Table III. Duration of pain, entered first into all regression equations, did not account for a significant proportion of the variance in the dependent variables studied: average level of pain, depression (as assessed by the Beck Depression Inventory), downtime, and physical and psychosocial impairment (as assessed by the Sickness Impact Profile). The 3 CSQ factors, entered as a group, did add a significant proportion of the variance in predicting
361
depression, downtime, and total SIP scores as well as psychosocial and physical impairment, but not average pain. Independent coping factor regression coefficients were examined to determine which factors were significantly related to which dependent variables. The first factor, Denial of Pain, was significantly and positively related to downtime. That is, subjects who reported greater use of denial and reinterpretation strategies spent more time reclining during the day and evening. The seond factor, Diverting Attention and Praying, was significantly and positively related to average pain. Patients high on the third factor, Helplessness, were more depressed and more impaired as measured by physical, psychosocial, and total SIP scores. Treatment-related changes in coping strategies Analysis of covariance was performed to determine whether the 3 experimental conditions differed in utilization of coping strategies at the end of treatment. In each case, pretreatment scores on the CSQ subscales were entered as covariates. Significant differences were found on 4 of the 7 CSQ subscales: diverting attention (F= 4.29, df= 2, 67, P < 0.018) catastrophizing (F= 3.89, df = 2, 70, P < 0.025), ignoring sensations (F= 3.10, df = 2, 69, P < 0.05), and coping self-statements (F = 7.02, df = 2, 68, P -C 0.002). Planned contrasts indicated that the cognitive-behavioral group had significantly higher diverting attention scores at the end of treatment than did the waiting list control group (t = 2.90, P < 0.005). Both the cognitive and operant groups had significantly lower catastrophizing scores (t = -2.49, P < 0.015, and t = -2.43, P -C 0.018) and higher coping self-statement scores (t = 3.71, P < 0.001, and t = 2.53, P < 0.014) relative to the control group. Finally, the operant group had significantly higher ignoring sensations scores than did the control group (t = 2.46, P < 0.016). In order to determine if there were significant changes within each condition over the period of the treatment, paired-comparison t tests were computed. The results revealed significant increases in the cognitive group in both coping self-statements (t = 3.16, P < 0.004) and diverting attention strategies (t = 3.66, P < 0.001). Both the cognitive and operant groups had significantly increased scores on ignoring pain (t = 2.10, P < 0.047, and t = 2.28, P -C0.03), and a decrease in catastrophizing (t = - 3.30, P < 0.003, and t = -2.66, P < 0.013). Patients in the waiting list condition reported significant decreases in the use of coping self-statements (t = 2.65, P < 0.016). Relationships between changes in coping strategies and pain and disability In order to determine the association between changes in CSQ subscales and changes in reported pain intensity and functional disability over the course of the treatments, Pearson correlations were computed. Increased use of praying and hoping strategies was significantly related to decreases in reported pain intensity (r = - 0.21, P -C 0.05). Decreased endorsement of catastrophizing strategies was also related significantly to decreases in pain intensity ratings (r = 0.32, P < O.Ol), as well as to decreases in total scores on the SIP (r = 0.28, P < 0.01) and the SIP psychosocial subscale (r = 0.36, P -C0.001).
Discussion In comparing findings of this study to those of Rosenstiel and Keefe [lY], several potentially important sample differences should be kept in mind. For example, our sample consisted of 36% female subjects, in contrast to Rosenstiel and Keefe’s [lY] 69% female subject sample. Only 7% of our sample anticipated or were currently receiving financial compensation for pain, in contrast to 46% of the Rosenstiel and Keefe [19] sample. Finally, the present sample was on the average 3 years older than that in the Rosenstiel and Keefe [lY] study. Despite these sample differences, the two studies found very similar factor structures for the CSQ. In both cases, 3 factors were obtained, accounting for similar proportions of the variance in item responses. Differences between the two analyses took two forms. First, there were slight changes in subscale loadings on the first factor. These differences do not seem to change the interpretation of this factor as primarily reflecting a tendency to ignore pain sensations. Second, the second and third factors were reversed. Although there were some differences between the two studies in subscale loadings on each factor: both studies found that one factor related to feelings of helplessness and the other to use of attention diversion techniques, as well as to praying or hoping the pain would improve. Coping styles were found to be associated with average pain, downtime, functional impairment, and depression. Our findings generally corroborate those of Rosenstiel and Keefe [lY]. who reported significant relationships between coping style and average pain, depression, state anxiety, and functional capacity, but not to downtime. Relationships between individual coping factors and dependent measures differed somewhat between the two studies: this may well reflect the differences in dependent measures as well as differences in factor structure and sample characteristics. Important similarities between results of the two studies include the significant positive relationship between catastrophizing/feeling unable to control pain and depression, and between diverting attention and pain intensity. That catastrophizing and believing one has little ability to control pain are related to depression is certainly not surprising, in light of the large literature [cf. 201 demonstrating the association between such negative cognitions and depression in non-pain populations. The finding of the present study that decreased catastrophizing following psychological treatment for chronic back pain is associated with decreased pain and physical and psychosocial disability due to pain suggests the usefulness of incorporating techniques targeting this variable into pain treatment programs. That catastrophizing decreased in both the cognitive and the operant treatments in this study may indicate that this change was a result of decreased disability due to other treatment factors, or may reflect the emphasis in both treatments that pain could be controlled (although by different means). The relationship between attention diversion and praying or hoping and pain problems is more puzzling. As Rosenstiel and Keefe [19] pointed out, attention diversion techniques have been found useful in decreasing pain in experimental pain studies, but may not be feasible for chronic pain problems. In support of this hypothesis, Rybstein-Blinchik [21] found that in a group of chronic pain patients.
363
training in reinterpretation strategies was more effective in decreasing reported pain and pain behaviors than was a distraction method. Further, a recent review [14] concludes that distraction techniques may be more helpful in alleviating mild as opposed to severe pain. It would appear, in summary, that it is probably not useful to incorporate training in attention diversion techniques in chronic pain treatment programs. However, such a conclusion is based on preliminary data and needs to be substantiated in further direct empirical tests. It should also be emphasized that ‘attention diversion’ techniques in this context refer to strategies such as counting numbers or mentally reciting poems or songs, not to engaging in activities in order to decrease thinking and worrying about the pain. The latter may be an effective coping strategy for many patients. That increased praying or hoping following treatment was associated with decreased pain ratings suggests that the positive relationship between the Diverting Attention and Praying factor and pain may be due to the ineffectiveness of distraction techniques, and not to the ineffectiveness of praying and hoping. Further research is needed to elucidate this issue. The results of this and previous studies suggest the need for further research into ways individuals cope with chronic pain. At least 3 directions are indicated: development of more refined methods to assess coping strategies, attention to developing and empirically testing more sophisticated conceptual formulations, and the elucidation of environmental variables (e.g., social support) that may facilitate more adaptive coping.
Acknowledgements This research was supported by Grant No. NS 19619 to Judith A. Turner from the National Institute of Neurological and Communicative Disorders and Stroke of the National Institutes of Health. The authors wish to thank Drs. Anne Rosenstiel and Francis Keefe for making available the Coping Strategy Questionnaire, and Larry West and Andrew Fry for assistance with the study.
References 1 Anderson, N.B., Lawrence, P.S. and Olson, T.W., Within-subject analysis of autogenic training and cognitive coping training in the treatment of tension headache pain, J. behav. Ther. exp. Psychiat., 12 (1981) 219-223. 2 Beck, A.T., Ward, C.H., Mendelson, M., Mock, J. and Erbaugh, J., An inventory for measuring depression, Arch. gen. Psychiat., 4 (1961) 53-63. 3 Beers, T.M. and Karoly, P., Cognitive strategies, expectancy, and coping style in the control of pain, J. consult. clin. Psychol., 47 (1979) 179-180. 4 Bergner, M., Bobbitt, R.A., Carter, W.B. and Gilson, B.S., The Sickness Impact Profile: development and final revision of a health status measure, Med. Care, 19 (1981) 787-805. 5 Billings, A.G. and Moos, R.H., The role of coping responses and social resources in attenuating the stress of life events, J. behav. Med., 4 (1981) 139-157. 6 Bradley, L.A. add Van der Heide, L.H., Pain-related correlates of MMPI profile subgroups among back pain patients, Hlth Psychol., 3 (1984) 157-174.
364
7 Clmes, J.F. and Barber. T.X.. C’ognitive strategies. experimenter modeling, and expectation in the attenuation of pain. J. abnorm. Psychol., 83 (1974) 356-363. 8 Copp, L.A.. The spectrum of suffering, Amer. J. Nurs.. 74 (1974) 491-495. 9 Engstrom. D.. Cognitive behavioral therapy methods in chronic pain treatment. In: J.J. Bonica, U. Lindhlom. A. Iggo, L.E. Jones and C. Benedetti (Eds.). Advances in Pain Research and Therapy, Vol. 5. Raven Press, New York. 1983. pp. 829-838. 10 Fordyce, W.E., Behavioral Methods for Chronic Pain and Illness. Moshy, St. Louis, MO. 1976. 11 Holroyd. K.A. and Andrasik, F.. Coping and the self-control of chronic tension headache. J. consult. clin. Psychol., 46 (1978) 303663045. 12 Kendall. P.C.. Williams. L.. Pechacek, T.F.. Graham, L.E.. Shisslak, C. and Herzoff, N.. Cognitive-hehavioral and patient education interventions in cardiac catheterization procedures: the Palo Alto medical psychology project, J. consult, clin. Psychol.. 47 (1979) 49-58. 13 Kremsdorf, R.B.. Kochanowicz. N.A. and Costell, S., Cognitive skills training versus EMG biofeedback in the treatment of tension headaches, Biofeedback Self-Reg., 6 (1981) 93-102. 14 McCaul. K.D. and Malott. J.M.. Distraction and coping with pain, Psychol. Bull., 95 (1984) 516-533. 15 Mitchell, K.R. and White. R.G.. Behavioral self-management: an application to the problem of migraine headaches. Behav. Ther., 8 (1977) 213-222. 16 Moore. J.E. and Chaney, E.F.. Outpatient group treatment of chronic pain: effects of spouse involvement, J. consult. clin. Psychol., 53 (1985) 3266334. 17 Peariin, L.I. and Schooler, C.. The structure of coping, J. Hlth Sot. Behav.. 19 (1978) 2-21. 18 Pickett. C. and Glum, G.A.. Comparative treatment strategies and their interaction with locus of control in the reduction of postsurgical pain and anxiety, J. consult. clin. Psychol.. 50 (1982) 439-441. 19 Rosenstiel. A.K. and Keefe, F.J.. The use of coping strategies in chronic low back pain patients: relationship to patient characteristics and current adjustment, Pain, 17 (1983) 33-44. 20 Rush, J. and Weissenburger, J.. Depression and suicide. In: R. Grieger and I.Z. Grieger (Eds.). Cognition and Emotional Disturbance, Human Sciences Press, New York. 1982, pp. 766109. 21 Ryhstein-Blinchik. E., Effects of different cognitive strategies on chronic pain experience, J. behav. Med., 2 (1979) 93-101. 22 Turner, J.A., Comparison of group progressive-relaxation training and cognitive-behavioral group therapy for chronic low hack pain, J. consult. clin. Psychol., 50 (1982) 757-765. 23 Wernick, R.L., Jaremko, M.E. and Taylor, P.W.. Pain management in severely burned adults: a test of stress inoculation, J. hehav. Med.. 4 (1981) 103-109.
355
PAI 00844
Strategies for Coping with Chronic Low Back Pain: Relationship to Pain and Disability Judith A. Turner *,** and Stephen Clancy * * Department of Psychiatry and Behavroral Sciences, RP-IO, and ** Department of Rehabilitation Medicine, Multidisciplinary Pain Center, University of Washington School of Medicine, Seattle, WA 98195 (U.S.A.) (Received
22 May 1985, revised received 15 July 1985, accepted
24 July 1985)
Summary Seventy-four chronic low back pain patients in a study assessing the effectiveness of group outpatient cognitive-behavioral and operant behavioral treatment completed the Coping Strategy Questionnaire (CSQ) and measures of pain, depression, and functional disability pre- and post-treatment. The previously reported factor structure of the CSQ was generally replicated, and significant associations were found between use of ignoring and reinterpretation strategies and downtime, between use of attention diversion strategies and pain intensity, and between tendency to catastrophize and physical and psychosocial impairment. Both treatments resulted in significant changes in types of coping strategies used to deal with pain. Increased use of praying and hoping strategies was significantly related to decreases in pain intensity. Decreased catastrophizing was also significantly related to decreases in pain intensity, as well as to decreases in physical and psychosocial impairment.
Introduction There is considerable evidence that the coping responses of individuals to stressors play an important role in adjustment to the stress [5,17]. The question of how people cope with pain, and especially, of what constitutes effective and optimal coping, is of interest to a large number of researchers and clinicians. Coping, in this context, refers to thoughts and behaviors people use to manage their pain or their
Correspondence
0304-3959/86/$03.50
and requests
for reprints
should be addressed
0 1986 Elsevier Science Publishers
to Dr. Turner.
B.V. (Biomedical
Division)
emotional reactions to the pain, so as to reduce emotional distress. It appears that patients with pain use a wide variety of such strategies [Xl, and a number of studies have demonstrated the impact of training in specific coping skills on pain, distress. and other aspects of pain problems. In experimental pain studies, training in the use of cognitive strategies for pain control has been shown to increase pain tolerance and threshold [3] and decrease pain report [7]. Studies in clinical settings have suggested the effectiveness of coping skills training in reducing pain and distress associated with medical and surgical procedures [12,18.23]. Further. coping skills training has been demonstrated in several controlled studies to decrease headache activity [1.13.15] and pain ratings [9.16,22] of chronic pain patients. Although wide variation in degree of physical and psychological disability among individuals with chronic pain has been noted [6], very little is known about how the ways in which people cope with chronic pain may relate to such impairment. Further, although many chronic pain treatment programs emphasize teaching cognitive and behavioral skills to increase patients’ ability to cope with and control pain, empirical data concerning the effects of such treatments on the use of coping skills, as well as concerning the effects of changes in coping strategy utilization on pain and disability, are lacking. In a first step towards examining these issues, Rosenstiel and Keefe [19] described a measure of the extent to which chronic low back pain patients use 1 behavioral and 6 cognitive coping strategies. Three factors of related coping strategies were identified, accounting for most of the variance in responses on this measure, the Coping Strategy Questionnaire (CSQ). Each of the 3 factors was found to be associated with measures of behavioral and emotional adjustment to pain. The present study sought to determine whether this same factor structure would be replicated in a different patient sample, thus indicating that types of coping styles (functionally related coping strategies) as measured by the CSQ could be reliably identified across patients with different demographic and pain characteristics. The present study had several additional objectives: (a) to determine whether Rosenstiel and Keefe’s [19] findings related to the association between CSQ factors and physical and psychological functioning would be replicated in a new sample, (b) to examine whether cognitive-behavioral and operant behavioral treatments, as compared with a waiting list control condition. produced changes in coping strategy utilization as assessed by the CSQ, and (c) to examine whether CSQ changes were associated with changes in pain and disability.
Methods Subjects Subjects were 74 chronic low back pain patients (47 males and 27 females) who completed 1 of 3 conditions in a treatment outcome study. Subjects were referred by physicians or self-referred following media publicity. Criteria for study entry included the following subject characteristics: persistent low back pain for at least 6 months. age between 20 and 65, and married or co-habiting. Exclusion criteria
351
included back surgery within the past year and history of or current cancer or cardiac disease. The mean age of subjects was 46.3 years (S.D. = 10.0). The mean duration of time since onset of first back pain problem was 14.5 years (S.D. = 11.6) and mean duration of current back pain problem was 6.8 years (S.D. = 8.4). Five subjects were receiving financial compensation or anticipated future legal action related to pain. Ten subjects had undergone surgery for back pain. Measures
The Coping Strategy Questionnaire (CSQ) [19] lists 42 strategies for coping with pain, including 6 different types of cognitive strategies (diverting attention, reinterpreting pain sensations, coping self-statements, ignoring pain sensations, praying or hoping, and catastrophizing) and a behavioral strategy (increasing activity level). On the CSQ, subjects rate on a scale from 0 (never) to 6 (always) how often they use each strategy when they experience pain. They also rate how much control they believe they have over pain, and how much they are able to decrease pain, on a scale from 0 (no control/cannot decrease it at all) to 6 (complete control/can decrease it completely). Subjects also completed a Pain Diary [lo] for 1 week, making hourly ratings of pain intensity on a O-10 scale, and recording type of activity and position (sitting, standing or walking, or reclining). For the purpose of data analyses, following the protocol of Rosenstiel and Keefe [19], average pain was defined as the average of 3 subject ratings: (1) verbally reported pain intensity at the time of assessment, rated on a O-10 scale with anchors of ‘no pain’ and ‘pain as bad as it can be,’ (2) the most severe pain in the past week, as recorded on pain diaries, and (3) the least severe pain in the past week, as recorded on pain diaries. Downtime was defined as the amount of time spent lying down or in a reclining position between the hours of 7 a.m. and 11 p.m., as recorded on the pain diaries. The Sickness Impact Profile (SIP) [4] was used as a measure of pain-related physical and psychosocial disability. This behavioral checklist consists of 136 items in 12 areas of health-related dysfunction and yields a total impairment index as well as summary scores of physical and psychosocial dysfunction. The Beck Depression Inventory (BDI) [2], a widely used 21-item self-report measure, was used to assess current severity of depression. Procedure
Subjects completed the CSQ, SIP, and BDI after telephone screening and before random assignment to 1 of 3 conditions: waiting list control, cognitive-behavioral therapy, and operant behavioral therapy. The Pain Diaries were completed over a 1 week period following completion of the other measures, and prior to beginning treatment (in the case of those assigned to a treatment condition). All measures were completed again following treatment or the end of the 8 week waiting list period. Both treatments were conducted in a group format, in 2 h sessions weekly for 8 weeks. Each group consisted of 5-10 subjects, and spouses attended half the sessions in the operant behavioral treatment. Five groups were conducted in each treatment
TABLt
1
MEANS AND STANDARD DEVIArIONS STRATEGY QUESTIONNAIRE Subscales
OF
RAI’INGS
ON
SUBSCALES
Mean
S.D.
1.74 u.94 3.64 2.32 1.94 1.24
1.25 0.94 1.Oh 0.98 1.15 1.11
Behurwrul cqq struregl 1. Increasing activity level
2.40
1.1X
Ef/emwness rutingv 1. Control over pain 2. Ability to decrease
3.30 3.07
1.19 1.14
Co~nrtrw qprng 1. 2. 3. 4. 5. 6.
OF THE
COPING
.rtrrr1qrav
Diverting attention Reinterpreting pain sensations Coptng self-statements Ignoring pain sensations Praying or hoping Catastrophizing
pain
condition, and each group was led by 1 of 5 experienced Ph.D. level clinical psychologists. With one exception, each therapist led both a cognitive and an operant behavioral group. The cognitive-behavioral treatment included training in systematic progressive muscle relaxation, imagery, covert assertion, and the identification and modification of distorted, maladaptive thoughts related to pain and stressful events. The operant behavioral treatment included a regular aerobic walking program, training subjects and spouses in the identification of pain behaviors, instructing subjects to decrease pain behaviors, instructing spouses to not reinforce pain behaviors and to reinforce well behaviors, and communication training.
Results Twenty-one subjects completed the waiting list, 24 the cognitive-behavioral, and 29 the operant behavioral treatment. As no significant differences were found on demographic and dependent measure variables. the group cohorts within conditions were collapsed for purposes of data analysis. Replicution of CSQ factor structure
Differences were observed in the frequency with which subjects used the different coping strategies, as shown in Table I. As was found by Rosenstiel and Keefe [19], subjects reported they rarely reinterpreted pain sensations as a coping strategy (mean = 0.94). Coping self-statements were most frequently reported (mean = 3.64), followed by ignoring pain sensations (mean = 2.32) and praying and hoping (mean = 1.94). Although the overall ability to control and decrease pain was rated as somewhat higher by these subjects than by Rosenstiel and Keefe’s [19] sample, the scores nonetheless were rather low (mean = 3.30 and 3.07 on a scale of O-6).
359 TABLE
II
PRINCIPAL Factor
COMPONENT Component
ANALYSIS
OF THE COPING
measures
Factor 1
Denial of pain Ignoring pain sensations Reinterpreting pain sensations Ability to decrease pain Diverting attention and praying Diverting attention Praying or hoping Increasing activity Helplessness Catastrophizing Ability to control pain Coping self-statements
STRATEGY
QUESTIONNAIRE
loadings 2
3
0.65 0.81 0.74
0.13 0.16 -0.17
- 0.04 0.43 - 0.25
0.16 - 0.21 0.32
0.81 0.75 0.67
-0.10 0.13 -0.12
0.11 0.46 0.05
0.13 -0.11 0.38
0.75 - 0.60 - 0.69
To determine how the different coping strategies were related to one another and to ratings of ability to control and decrease pain, a principal components analysis, using oblique rotation, was performed. Of the factors emerging from this analysis, only those having eigenvalues of 1 or greater were considered. A strategy was included in a factor if (1) it correlated with the factor at a level greater than 0.50, and (2) it had its highest loading on that factor. This analysis produced 3 factors, as shown in Table II, that accounted for 65% of the variance in CSQ responses. This solution was similar to that in Rosenstiel and Keefe’s [19] study, which also found a 3-factor solution, accounting for 68% of the variance. The first factor, accounting for 33% of the variance, was very similar to the first factor reported by Rosenstiel and Keefe [19], which they labeled Cognitive Coping and Suppression. Individuals high on this factor reported that they attempted to suppress pain through ignoring and reinterpreting pain sensations. They also rated themselves as able to some extent to decrease pain, a variable with a very low loading on the Rosenstiel and Keefe [19] component analysis. Unlike the results found by Rosenstiel and Keefe [19], coping self-statements did not load highly on this factor, which we labeled ‘denial of pain.’ The second factor, accounting for 20% of the variance, was similar to the third factor reported by Rosenstiel and Keefe [19]. Individuals high on this factor reported diverting attention, praying and hoping, and increasing activity as methods used to reduce pain. As did Rosenstiel and Keefe [19], we labeled this factor ‘diverting attention and praying.’ The third factor, similar to the factor labeled Helplessness by Rosenstiel and Keefe [19], accounted for 12% of the variance. Individuals high on this factor, which we also call ‘helplessness,’ frequently endorsed items on the catastrophizing subscale and rated their effectiveness in controlling pain as low. In addition, subjects high on this factor rarely used coping self-statements.
TABLE
III
REGRESSION ANALYSES: LOGICAL AND PHYSICAL Measure
of adjustment
I. Average pain
Denial of pain Diverting attention Helplessness 2. Depresston Denial of pain Diverting attention Helplessness
RELATIONSHIP DISABILITY Total R
0.33
and praying
OF COPING
F ratio for R change
F(3.68)
=
2.10
F(l,68)
=
0.17
F(1.
68) =
F(1.68) 0.54
and praying
F(3.
0.53
4. Physical impairment
0.40
=
68) =
TO PAIN AND
Proportion acct. for by 3 factors (S)
6.91 *** 0.36
F(1.68)
=
0.00
F(3,68)
=
6.44 **
=
1.28
F(3.68)
= 3.22 **
R
29 - 0.04 0.05 0.53
68) = 27.02 ***
F(1.68) F(1,68) F(l.68)
Simple
0.04 0.33 - 0.01
0.10
=
PSYCHO-
11
8.28 **
F(l,68) F(1.
3. Dtsability (SIP total) Denial of pain Diverting attention and praying Helplessness
FACTORS
27 0.07 0.23 0.47
= 2.38 = 20.71 *** 16
(SIP physical)
Denial of pain Diverting attention Helplesness
5. Psychosoctal impairment
0.12 0.21 0.32
F(1, 68) = 1.56 F(1,68)= 1.79 F(1, 68) = 8.84 **
and praying
0.52
F(3, 68) = 6.20 ***
27
(SIPpsychosocml)
Denial of pain Diverting attention Helplessness 6. Downtrme Denial of pain Diverting attention Helplessness
0.41 and praying
0.07 0.18 0.48
F(1, 68) = 1.72 F(l, 68) = 1.12 F(1, 68) = 21.78 ***
and praying
F(3,68) F(1, 68) F(1, 68) F(1.68)
= = = =
3.35 ** 6.14 * 2.14 0.75
16 0.31 0.27 0.06
* P < 0.02. ** PC
0.01. *** P < 0.001.
Relationship between coping strategies and physical and psychological functioning We next performed hierarchical regression analyses to examine the relationship between coping strategies and adjustment to pain, after accounting for the effect of duration of pain. The results are shown in Table III. Duration of pain, entered first into all regression equations, did not account for a significant proportion of the variance in the dependent variables studied: average level of pain, depression (as assessed by the Beck Depression Inventory), downtime, and physical and psychosocial impairment (as assessed by the Sickness Impact Profile). The 3 CSQ factors, entered as a group, did add a significant proportion of the variance in predicting
361
depression, downtime, and total SIP scores as well as psychosocial and physical impairment, but not average pain. Independent coping factor regression coefficients were examined to determine which factors were significantly related to which dependent variables. The first factor, Denial of Pain, was significantly and positively related to downtime. That is, subjects who reported greater use of denial and reinterpretation strategies spent more time reclining during the day and evening. The seond factor, Diverting Attention and Praying, was significantly and positively related to average pain. Patients high on the third factor, Helplessness, were more depressed and more impaired as measured by physical, psychosocial, and total SIP scores. Treatment-related changes in coping strategies Analysis of covariance was performed to determine whether the 3 experimental conditions differed in utilization of coping strategies at the end of treatment. In each case, pretreatment scores on the CSQ subscales were entered as covariates. Significant differences were found on 4 of the 7 CSQ subscales: diverting attention (F= 4.29, df= 2, 67, P < 0.018) catastrophizing (F= 3.89, df = 2, 70, P < 0.025), ignoring sensations (F= 3.10, df = 2, 69, P < 0.05), and coping self-statements (F = 7.02, df = 2, 68, P -C 0.002). Planned contrasts indicated that the cognitive-behavioral group had significantly higher diverting attention scores at the end of treatment than did the waiting list control group (t = 2.90, P < 0.005). Both the cognitive and operant groups had significantly lower catastrophizing scores (t = -2.49, P < 0.015, and t = -2.43, P -C 0.018) and higher coping self-statement scores (t = 3.71, P < 0.001, and t = 2.53, P < 0.014) relative to the control group. Finally, the operant group had significantly higher ignoring sensations scores than did the control group (t = 2.46, P < 0.016). In order to determine if there were significant changes within each condition over the period of the treatment, paired-comparison t tests were computed. The results revealed significant increases in the cognitive group in both coping self-statements (t = 3.16, P < 0.004) and diverting attention strategies (t = 3.66, P < 0.001). Both the cognitive and operant groups had significantly increased scores on ignoring pain (t = 2.10, P < 0.047, and t = 2.28, P -C0.03), and a decrease in catastrophizing (t = - 3.30, P < 0.003, and t = -2.66, P < 0.013). Patients in the waiting list condition reported significant decreases in the use of coping self-statements (t = 2.65, P < 0.016). Relationships between changes in coping strategies and pain and disability In order to determine the association between changes in CSQ subscales and changes in reported pain intensity and functional disability over the course of the treatments, Pearson correlations were computed. Increased use of praying and hoping strategies was significantly related to decreases in reported pain intensity (r = - 0.21, P -C 0.05). Decreased endorsement of catastrophizing strategies was also related significantly to decreases in pain intensity ratings (r = 0.32, P < O.Ol), as well as to decreases in total scores on the SIP (r = 0.28, P < 0.01) and the SIP psychosocial subscale (r = 0.36, P -C0.001).
Discussion In comparing findings of this study to those of Rosenstiel and Keefe [lY], several potentially important sample differences should be kept in mind. For example, our sample consisted of 36% female subjects, in contrast to Rosenstiel and Keefe’s [lY] 69% female subject sample. Only 7% of our sample anticipated or were currently receiving financial compensation for pain, in contrast to 46% of the Rosenstiel and Keefe [19] sample. Finally, the present sample was on the average 3 years older than that in the Rosenstiel and Keefe [lY] study. Despite these sample differences, the two studies found very similar factor structures for the CSQ. In both cases, 3 factors were obtained, accounting for similar proportions of the variance in item responses. Differences between the two analyses took two forms. First, there were slight changes in subscale loadings on the first factor. These differences do not seem to change the interpretation of this factor as primarily reflecting a tendency to ignore pain sensations. Second, the second and third factors were reversed. Although there were some differences between the two studies in subscale loadings on each factor: both studies found that one factor related to feelings of helplessness and the other to use of attention diversion techniques, as well as to praying or hoping the pain would improve. Coping styles were found to be associated with average pain, downtime, functional impairment, and depression. Our findings generally corroborate those of Rosenstiel and Keefe [lY]. who reported significant relationships between coping style and average pain, depression, state anxiety, and functional capacity, but not to downtime. Relationships between individual coping factors and dependent measures differed somewhat between the two studies: this may well reflect the differences in dependent measures as well as differences in factor structure and sample characteristics. Important similarities between results of the two studies include the significant positive relationship between catastrophizing/feeling unable to control pain and depression, and between diverting attention and pain intensity. That catastrophizing and believing one has little ability to control pain are related to depression is certainly not surprising, in light of the large literature [cf. 201 demonstrating the association between such negative cognitions and depression in non-pain populations. The finding of the present study that decreased catastrophizing following psychological treatment for chronic back pain is associated with decreased pain and physical and psychosocial disability due to pain suggests the usefulness of incorporating techniques targeting this variable into pain treatment programs. That catastrophizing decreased in both the cognitive and the operant treatments in this study may indicate that this change was a result of decreased disability due to other treatment factors, or may reflect the emphasis in both treatments that pain could be controlled (although by different means). The relationship between attention diversion and praying or hoping and pain problems is more puzzling. As Rosenstiel and Keefe [19] pointed out, attention diversion techniques have been found useful in decreasing pain in experimental pain studies, but may not be feasible for chronic pain problems. In support of this hypothesis, Rybstein-Blinchik [21] found that in a group of chronic pain patients.
363
training in reinterpretation strategies was more effective in decreasing reported pain and pain behaviors than was a distraction method. Further, a recent review [14] concludes that distraction techniques may be more helpful in alleviating mild as opposed to severe pain. It would appear, in summary, that it is probably not useful to incorporate training in attention diversion techniques in chronic pain treatment programs. However, such a conclusion is based on preliminary data and needs to be substantiated in further direct empirical tests. It should also be emphasized that ‘attention diversion’ techniques in this context refer to strategies such as counting numbers or mentally reciting poems or songs, not to engaging in activities in order to decrease thinking and worrying about the pain. The latter may be an effective coping strategy for many patients. That increased praying or hoping following treatment was associated with decreased pain ratings suggests that the positive relationship between the Diverting Attention and Praying factor and pain may be due to the ineffectiveness of distraction techniques, and not to the ineffectiveness of praying and hoping. Further research is needed to elucidate this issue. The results of this and previous studies suggest the need for further research into ways individuals cope with chronic pain. At least 3 directions are indicated: development of more refined methods to assess coping strategies, attention to developing and empirically testing more sophisticated conceptual formulations, and the elucidation of environmental variables (e.g., social support) that may facilitate more adaptive coping.
Acknowledgements This research was supported by Grant No. NS 19619 to Judith A. Turner from the National Institute of Neurological and Communicative Disorders and Stroke of the National Institutes of Health. The authors wish to thank Drs. Anne Rosenstiel and Francis Keefe for making available the Coping Strategy Questionnaire, and Larry West and Andrew Fry for assistance with the study.
References 1 Anderson, N.B., Lawrence, P.S. and Olson, T.W., Within-subject analysis of autogenic training and cognitive coping training in the treatment of tension headache pain, J. behav. Ther. exp. Psychiat., 12 (1981) 219-223. 2 Beck, A.T., Ward, C.H., Mendelson, M., Mock, J. and Erbaugh, J., An inventory for measuring depression, Arch. gen. Psychiat., 4 (1961) 53-63. 3 Beers, T.M. and Karoly, P., Cognitive strategies, expectancy, and coping style in the control of pain, J. consult. clin. Psychol., 47 (1979) 179-180. 4 Bergner, M., Bobbitt, R.A., Carter, W.B. and Gilson, B.S., The Sickness Impact Profile: development and final revision of a health status measure, Med. Care, 19 (1981) 787-805. 5 Billings, A.G. and Moos, R.H., The role of coping responses and social resources in attenuating the stress of life events, J. behav. Med., 4 (1981) 139-157. 6 Bradley, L.A. add Van der Heide, L.H., Pain-related correlates of MMPI profile subgroups among back pain patients, Hlth Psychol., 3 (1984) 157-174.
364
7 Clmes, J.F. and Barber. T.X.. C’ognitive strategies. experimenter modeling, and expectation in the attenuation of pain. J. abnorm. Psychol., 83 (1974) 356-363. 8 Copp, L.A.. The spectrum of suffering, Amer. J. Nurs.. 74 (1974) 491-495. 9 Engstrom. D.. Cognitive behavioral therapy methods in chronic pain treatment. In: J.J. Bonica, U. Lindhlom. A. Iggo, L.E. Jones and C. Benedetti (Eds.). Advances in Pain Research and Therapy, Vol. 5. Raven Press, New York. 1983. pp. 829-838. 10 Fordyce, W.E., Behavioral Methods for Chronic Pain and Illness. Moshy, St. Louis, MO. 1976. 11 Holroyd. K.A. and Andrasik, F.. Coping and the self-control of chronic tension headache. J. consult. clin. Psychol., 46 (1978) 303663045. 12 Kendall. P.C.. Williams. L.. Pechacek, T.F.. Graham, L.E.. Shisslak, C. and Herzoff, N.. Cognitive-hehavioral and patient education interventions in cardiac catheterization procedures: the Palo Alto medical psychology project, J. consult, clin. Psychol.. 47 (1979) 49-58. 13 Kremsdorf, R.B.. Kochanowicz. N.A. and Costell, S., Cognitive skills training versus EMG biofeedback in the treatment of tension headaches, Biofeedback Self-Reg., 6 (1981) 93-102. 14 McCaul. K.D. and Malott. J.M.. Distraction and coping with pain, Psychol. Bull., 95 (1984) 516-533. 15 Mitchell, K.R. and White. R.G.. Behavioral self-management: an application to the problem of migraine headaches. Behav. Ther., 8 (1977) 213-222. 16 Moore. J.E. and Chaney, E.F.. Outpatient group treatment of chronic pain: effects of spouse involvement, J. consult. clin. Psychol., 53 (1985) 3266334. 17 Peariin, L.I. and Schooler, C.. The structure of coping, J. Hlth Sot. Behav.. 19 (1978) 2-21. 18 Pickett. C. and Glum, G.A.. Comparative treatment strategies and their interaction with locus of control in the reduction of postsurgical pain and anxiety, J. consult. clin. Psychol.. 50 (1982) 439-441. 19 Rosenstiel. A.K. and Keefe, F.J.. The use of coping strategies in chronic low back pain patients: relationship to patient characteristics and current adjustment, Pain, 17 (1983) 33-44. 20 Rush, J. and Weissenburger, J.. Depression and suicide. In: R. Grieger and I.Z. Grieger (Eds.). Cognition and Emotional Disturbance, Human Sciences Press, New York. 1982, pp. 766109. 21 Ryhstein-Blinchik. E., Effects of different cognitive strategies on chronic pain experience, J. behav. Med., 2 (1979) 93-101. 22 Turner, J.A., Comparison of group progressive-relaxation training and cognitive-behavioral group therapy for chronic low hack pain, J. consult. clin. Psychol., 50 (1982) 757-765. 23 Wernick, R.L., Jaremko, M.E. and Taylor, P.W.. Pain management in severely burned adults: a test of stress inoculation, J. hehav. Med.. 4 (1981) 103-109.