OUP user menu

Pain Catastrophizing and Pain Coping among Methadone-Maintained Patients

Brian Garnet BA, Mark Beitel PhD, Christopher J. Cutter PhD, Jonathan Savant BS, Skye Peters BA, Richard S. Schottenfeld MD, Declan T. Barry PhD
DOI: http://dx.doi.org/10.1111/j.1526-4637.2010.01002.x 79-86 First published online: 1 January 2011

Abstract

Objective. The aim of this study was to examine the association of pain catastrophizing and pain coping strategies with characteristic pain intensity (an average of worst, least, and typical pain intensity in the past week) and recent pain-related disability (an average of three measures of past week pain interference) in opioid-dependent patients enrolled in a methadone maintenance treatment program (MMTP) who reported recent pain.

Design. Cross-sectional survey.

Patients. One hundred and eight MMTP patients who reported recent pain.

Measures. Participants completed measures of demographics, pain status (i.e., “chronic severe pain”[pain lasting at least 6 months with at least moderate pain intensity or significant pain interference in the past week] vs “some pain”[pain in the past week not meeting the threshold of chronic severe pain]), characteristic pain intensity, recent pain-related disability, somatization, depression, catastrophizing, and pain coping strategies.

Results. Catastrophizing explained a significant proportion of the variance in characteristic pain intensity (14%) and recent pain-related disability (11%) after controlling for demographics, pain status, somatization, and depression. Mirroring the findings of studies of non-opioid-dependent chronic pain patients, greater catastrophizing was associated with greater pain intensity and increases in recent pain-related disability. On average, the chronic severe pain group reported higher levels of catastrophizing than the some pain group.

Conclusion. Consistent with studies of patients with chronic pain who are not opioid dependent, our findings emphasize the importance of assessing and addressing catastrophizing in MMTP patients with pain.

  • Pain
  • Opioid Dependence
  • Catastrophizing
  • Coping
  • Methadone

Introduction

Surprisingly little research has been conducted on pain coping strategies among opioid-dependent patients1 in methadone maintenance treatment programs (MMTPs), where pain prevalence estimates are high and range from 37% with chronic severe pain (CSP) (i.e., pain lasting at least 6 months with at least moderate pain intensity or significant pain interference in the past week) to more than 60% with chronic pain of any intensity [1–5]. Chronic pain in opioid-dependent patients attending MMTPs is associated with increased psychopathology, including elevated levels of anxiety, depression, personality disorder criteria, histories of attempted suicide, trauma, and functional disability [1,2,4,6].

According to Lazarus and Folkman, coping is a function of individuals' appraisal of events, and the use of coping strategies is based on judgments of perceived threat and perceived resources [7]. Pain coping refers to efforts made by individuals with pain to minimize, tolerate, or reduce their pain [8]. Coping strategies and coping skills training have been emphasized in pain management and opioid dependence treatment settings, respectively [9–14]. The Coping Strategies Questionnaire (CSQ [14]) and CSQ-Revised (CSQ-R [15]), which assess passive (i.e., praying and hoping in response to pain) and active (i.e., diverting attention, ignoring pain, reinterpreting pain sensations, and making coping self-statements) cognitive and behavioral pain coping strategies as well as catastrophizing, a maladaptive response to the experience of pain involving helplessness and pessimistic cognitions (e.g., “It's awful and I feel that it overwhelms me”), have been widely used in studies on chronic pain patients without substance use disorders. These studies have generally found that 1) catastrophizing is strongly associated with increased depression, physical and psychosocial impairment, and disability [9,14,16–18]; 2) passive coping strategies are ineffective in reducing pain intensity; and 3) active strategies are associated with both adaptive (i.e., decreased pain intensity and depression) and maladaptive (i.e., greater physical disability) outcomes [9,14,17,19,20]. Patients in MMTPs with co-occurring chronic pain and opioid dependence are a heterogeneous, complex, understudied group. For example, the onset of opioid dependence precedes that of chronic pain for some proportion of these patients, whereas the onset of chronic pain precedes that of opioid dependence for others. Specifically, remove “addiction” and replace it with “dependence,” and then switch the placement of the second “dependence” with the placement of the second “chronic pain.” The extent to which findings on catastrophizing and pain coping strategies among non-addicted chronic pain patients generalize to those with co-occurring chronic pain and opioid dependence in MMTPs is unclear, and this information may be useful to MMTPs as they consider developing pain management programs.

While we previously reported on the prevalence of pain types, their associated substance use and psychiatric correlates, and use of conventional and nonconventional pain treatments among the study sample [1,21], we did not report on patients' pain catastrophizing, pain coping strategies, characteristic pain intensity, or recent pain-related disability. Consequently, the primary goal of this study was to examine the association of pain catastrophizing and pain coping strategies with characteristic pain intensity and recent pain-related disability in MMTP patients reporting either CSP or “some pain” (SP), which is pain within the past 7 days not meeting the threshold of CSP. Prior research has linked sex, age, employment, pain status, somatization, and depression with pain intensity or pain disability [16,18,22–24]. Consequently, we hypothesized that, after controlling for demographics (sex, age, and employment status), pain status, somatization, and depression, higher levels of catastrophizing would be associated with increased characteristic pain intensity and recent pain-related disability. As a previous study found that the CSP group, on average, was older and exhibited higher levels of somatization than the SP group [1], we hypothesized that, after controlling for age and somatization, patients in the CSP group would report higher levels of catastrophizing than those in the SP group.

Methods

Participants and Procedures

Participants in this study were drawn from a larger study of patients enrolled in a MMTP. Detailed information regarding patient characteristics for the entire sample (N = 150) has been reported elsewhere [1]. In the current study, we included only patients who reported pain in the past 7 days (N = 115) and who completed the CSQ-R. Of the 115 patients who reported pain in the past 7 days, the majority, but not all, were administered the CSQ-R (92%, or 108/115) due to an error in the survey administration instructions. Characteristics for this sample of 108 participants are reported in Table 1.

View this table:
Table 1

Demographic and treatment-related characteristics for patients with at least some pain in the past 7 days (N = 108)

Sex, %
  Women44
  Men56
Age, mean (SD), years42.2 (9.8)
Race/ethnicity, %
  American Indian or Native Alaskan3
  Black32
  Hispanic9
  White56
Marital status, %
  Never married50
  Married or cohabitant14
  Divorced, separated, or widowed36
Level of education, %
  <high school33
  High school or GED34
  Some college or some vocational training22
  Associate, bachelor degree, or vocational license11
Employment status, %
  At least some employment20
  Not working, but not disabled46
  Disabled34
Methadone dose, mean (SD), mg90.5 (28.9)
Duration of treatment at MMTP, mean (SD), months49.4 (57.9)
Number of different MMTP enrollments, mean (SD)2.2 (1.8)
  • SD = standard deviation; GED = General Equivalency Diploma; MMTP = methadone maintenance treatment program.

Study participants were recruited at the Legion, Park, and Orchard MMTPs of the APT Foundation, a nonprofit organization with a census of approximately 1,500 at the time of data collection. While the APT Foundation has a primary care clinic that offers routine and specialty medical care (e.g., HIV, hepatitis), it did not provide specialty pain diagnostic or treatment services at the time when this survey was administered. Patients were self-selected in response to study fliers posted in each of the clinics asking for volunteers to complete a survey about their experiences as a patient at the APT Foundation. Patients were recruited between March 2007 and March 2008. All patients who spoke with a research assistant agreed to participate and completed the survey. Participants were blind to the specific aims of the study and were interviewed regardless of their pain status. Research assistants administered the questionnaire packet (measures described below) after describing the study, including potential risks and benefits of study participation. Participants were compensated $10 for study participation. This study, involving the use of survey data without identifiers, was presented to the Human Investigations Committee at APT and the Yale University School of Medicine, and was exempted from review per U.S. Department of Health and Human Services regulation 45 CFR 6.101(b) [2].

Measures

Demographic Information

The survey included items concerning patients' sex, age, race/ethnicity, marital or relationship status, level of education, employment status, and treatment characteristics (methadone dose, duration of treatment at MMTP, number of different MMTP enrollments).

Characteristic Pain Intensity, Recent Pain-Related Disability, and Pain Status

Average pain intensity was measured by asking patients to rate their pain intensity on an 11-point scale (0–10) adapted from the Brief Pain Inventory (BPI [25,26]) for “average pain in the past 7 days,”“worst pain in the past 7 days,” and “current pain.” The mean of these three intensity ratings × 10 was calculated to find the characteristic pain intensity following the scoring system from the Graded Chronic Pain Scale (GCPS [27,28]). In addition, three pain interference items (scored from 0–10) adapted from the BPI assessed the extent to which pain in the last 7 days had interfered with patients' “everyday life,”“normal work or activities,” and “relationships with other people.” The mean of these three interference ratings × 10 was calculated to find the recent pain-related disability score following the scoring system from the GCPS. Respondents were categorized according to their pain status (i.e., CSP vs SP). As previously done, CSP was defined as pain lasting at least 6 months with a rating of at least 5 on the worst pain intensity in the past 7 days or a rating of at least 5 on any of the pain interference items, and SP was defined as pain in the past 7 days not meeting the threshold of CSP [1].

CSQ-R

The CSQ-R [15] is a 27-item revision of the 42-item CSQ [14]. It assesses five cognitive and behavioral pain coping strategies: 1) diverting attention (e.g., “I do things I enjoy, such as watching TV or listening to music”); 2) ignoring sensations (e.g., “I ignore it”); 3) reinterpreting pain sensations (e.g., “I imagine that the pain is outside of my body”); 4) coping self-statements (e.g., “I tell myself to be brave and to carry on despite the pain”); and 5) praying–hoping (e.g., “I pray for the pain to stop”), as well as pain-related catastrophizing (e.g., “It's terrible and I feel it's never going to get any better”). Respondents rate the frequency of their use of the five specific coping strategies and catastrophizing when experiencing pain on a 7-point scale from 0 (“I never do that”) to 6 (“I always do that”). CSQ-R subscales have demonstrated adequate internal consistency and validity in a variety of clinical populations and healthy subjects [8,19,22,29,30]. In the current study, the internal consistency (Cronbach alpha coefficients) of the reinterpreting pain sensations, catastrophizing, and praying–hoping CSQ-R subscales ranged from 0.81 to 0.86, and the Cronbach alphas of the diverting attention, ignoring sensations, and coping self-statements subscales ranged from 0.68 to 0.76.

Brief Symptoms Inventory 18

The Brief Symptoms Inventory 18 (BSI-18 [31]) was used to screen for somatization and depression symptoms. Patients rated how bothered they were by symptoms during the past 7 days using a 5-point Likert-type scale ranging from 0 (“not at all”) to 4 (“extremely”). Raw scores were converted to area T scores (M = 50, standard deviation [SD] = 10) to facilitate interpretation. T scores ≥63 (90th percentile) are demonstrated to represent clinically significant psychiatric symptoms [31]. In this study, participants' raw scores were converted to T scores using the BSI-18 community sample norms [31]. The BSI-18 has been utilized in studies with a variety of community and medical samples, including those with pain and substance-related disorders [32–34]. In the current study, the Cronbach alpha coefficients of the depression and somatization BSI-18 subscales were 0.89 and 0.77, respectively.

Data Analyses

We examined the distributions of all study variables to ensure that they met the normal distribution assumptions for the statistical tests performed. Two subscales of the CSQ-R, reinterpreting pain sensations and praying–hoping, were significantly skewed positively and negatively, respectively; therefore, we performed a square root transformation of these variables in all data analyses.

We ran two regression analyses involving the association of the CSQ-R subscales with characteristic pain intensity and recent pain-related disability, respectively. We controlled for the following characteristics in the regression models as previous research has linked them to pain intensity or pain disability: sex, age, employment, pain status, somatization, and depression [16,18,22–24].

While Pearson and Spearman's rho correlations do not calculate independent contributions of specific variables (as they ignore the effects of other variables), multiple regression analyses provide two types of correlation that do calculate independent contributions: semi-partial and partial correlations. When the correlation between a specific predictor (e.g., catastrophizing) and the criterion (e.g., recent pain-related disability) partials out all of the remaining predictor variables from the relevant category of predictor variables (e.g., characteristic pain intensity), it is called a semi-partial correlation; when the correlation partials out the remaining predictor variables from both the specific predictor and the criterion variable, the correlation is called a partial correlation. Catastrophizing and characteristic pain intensity were entered as independent variables, and pain-related disability was entered as the dependent variability. We performed separate semi-partial and partial correlational analyses to examine the association between catastrophizing and recent pain-related disability, controlling for characteristic pain intensity.

Pain group (i.e., CSP vs SP) differences on the CSQ-R coping subscales were tested using a multivariate analysis of variance followed by multivariate analyses of covariance, controlling for age and somatization, which we previously found differed between the CSP and SP groups [1]. All analyses were performed using SPSS 15 (SPSS, Inc., Chicago, IL), and significance was set at the 0.05 level.

Results

Demographics

As summarized in Table 1, of the 108 participants (60 men and 48 women) aged 21–61 years (mean 42.2, SD 9.8), the majority was white (56%), unemployed (46%) or disabled (34%), and one-half never married. A majority had at least a high school level of education (67%). On average, respondents were attending their MMTP for 49.4 months (SD 57.9) and were receiving 90.5 mg of methadone daily (SD 28.9). Race/ethnicity was not significantly associated with pain catastrophizing, characteristic pain intensity, or recent pain-related disability in the current study.

Characteristic Pain Intensity, Recent Pain-Related Disability, and Pain Coping

Of the 108 patients reporting CSP (N = 56) or SP (N = 52), characteristic pain intensity ranged from 13 to 100 with a mean of 63 (SD = 22), and recent pain-related disability ranged from 0 to 100 with a mean of 44 (SD = 30). The mean (and SD) scores for the CSQ-R subscales stratified by pain group are summarized in Table 3. Generally, the most frequently endorsed coping strategies among the CSP and SP groups were praying–hoping and coping self-statements, while the least frequently endorsed was reinterpreting pain sensations.

View this table:
Table 3

Comparison of SP and CSP groups on a measure of coping strategies

SP (N = 52)CSP (N = 56)MANOVA (df, 1, 106)MANCOVA* (df, 1, 104)
MeasureMSDMSDFPFP
CSQ-R (0–6 scale)
  Diverting attention2.51.32.91.22.700.1040.010.925
  Catastrophizing1.81.32.71.213.64<0.0014.930.029
  Ignoring sensations2.21.21.81.22.850.0952.740.101
  Reinterpreting pain sensations1.51.41.21.31.700.1955.740.018
  Coping self-statements3.51.43.71.10.890.3490.550.459
  Praying–hoping3.52.14.41.53.630.0590.470.494
  • * Controlling for age and somatization.

  • SP = some pain in the past 7 days; CSP = current chronic severe pain; MANOVA = multivariate analysis of variance; MANCOVA = multivariate analysis of covariance; CSQ-R = Coping Strategies Questionnaire-Revised; SD = standard deviation.

Association of Somatization, Depression, Coping Strategies, and Catastrophizing with Characteristic Pain Intensity and Recent Pain-Related Disability

The results of the regression analyses are shown in Table 2. In the first model that examined associations between the CSQ-R subscales and characteristic pain intensity, all three steps accounted for significant proportions of the variance after adjusting for previous steps. Demographics/pain status accounted for 31% (P < 0.001), with significant and independent contributions from both employment status (P = 0.010) and pain status (P < 0.001), but not sex (P = 0.259) or age (P = 0.679). Somatization and depression accounted for an additional 11% (P < 0.001) with a significant and independent contribution from somatization (P < 0.001), but not depression (P = 0.056). The CSQ-R subscale scores accounted for an additional 14% (P < 0.001) of the variance with a significant and independent contribution from catastrophizing (P < 0.001).

View this table:
Table 2

Association of demographic, pain status, somatization, depression, coping, and catastrophizing measures with pain intensity and pain-related disability (N = 108)

MeasureR2ΔR2ΔFPβP
Characteristic pain intensity
  Demographics/pain status0.310.3111.67<0.001
    Sex−0.100.259
    Age0.040.679
    Employment status0.240.010
    Pain status0.40<0.001
  BSI-180.420.119.18<0.001
    Somatization0.44<0.001
    Depression−0.180.056
  Coping/catastrophizing0.550.144.85<0.001
    Diverting attention−0.200.057
    Catastrophizing0.38<0.001
    Ignoring sensations0.020.834
    Reinterpreting pain sensations0.080.426
    Coping self-statements0.050.598
    Praying–hoping0.170.053
Recent pain-related disability
  Demographics/pain status0.310.3111.17<0.001
    Sex−0.040.617
    Age−0.060.500
    Employment status0.36<0.001
    Pain status0.320.001
  BSI-180.390.086.500.002
    Somatization0.290.007
    Depression0.050.599
  Coping/catastrophizing0.490.113.320.005
    Diverting attention−0.040.750
    Catastrophizing0.34<0.001
    Ignoring sensations−0.030.809
    Reinterpreting pain sensations0.080.467
    Coping self-statements−0.110.229
    Praying–hoping−0.030.728
  • BSI = Brief Symptoms Inventory.

In the model predicting recent pain-related disability, all three steps of the model accounted for significant proportions of variance after adjusting for previous steps. Demographics/pain status accounted for 31% (P < 0.001) with significant and independent contributions from both employment status (P < 0.001) and pain status (P = 0.001). The psychiatric variables accounted for an additional 8% (P = 0.002) with a significant and independent contribution from somatization (P = 0.007), but not depression (P = 0.599). The CSQ-R subscales accounted for an additional 11% (P = 0.005) of the variance with a significant and independent contribution from catastrophizing (P < 0.001).

Catastrophizing and characteristic pain intensity were jointly related to recent pain-related disability (adjusted R2 = 0.59, F2,105 = 79.26, P < 0.001). Characteristic pain intensity (β = 0.63, t = 8.73, P < 0.001) and catastrophizing (β = 0.24, t = 3.28, P < 0.01) independently predicted pain-related disability. Catastrophizing continued to predict pain-related disability after controlling for characteristic pain intensity through partial (rpartial = 0.31, P < 0.05) and semi-partial correlation (rsemi-partial = 0.20, P < 0.05).

Coping Strategies and Catastrophizing Stratified by Pain Status

As summarized in Table 3, in comparison to the SP group, the CSP group reported statistically significant higher levels of catastrophizing. After controlling for age and somatization, group differences on the reinterpreting pain sensations subscale reached statistical significance: The CSP group reported lower levels of reinterpreting pain sensations and higher levels of catastrophizing than the SP group.

Discussion

To our knowledge, this study is among the first to systematically assess pain catastrophizing and pain coping strategies among opioid-dependent patients in a MMTP. Overall, we found strong support for our hypotheses that 1) catastrophizing would be associated with increased characteristic pain intensity and recent pain-related disability after controlling for demographics, pain status, somatization, and depression; and 2) the CSP group would report higher levels of catastrophizing than the SP group after controlling for age and somatization.

Our findings on the positive associations among catastrophizing, characteristic pain intensity, and recent pain-related disability among opioid-dependent patients with SP or CSP extend those previously documented in studies of chronic pain patients without substance use disorders, where catastrophizing has been consistently associated with increased psychological and functional disability [9,16,18,19,22,35]. While the relationships between catastrophizing and pain, and catastrophizing and depression have long been emphasized in the literature [9,36], we found that when both depression and catastrophizing were entered as independent predictors into a regression model, catastrophizing was strongly associated with both characteristic pain intensity and recent pain-related disability, while depression was not. Additionally, in the multivariable analysis, catastrophizing continued to independently predict recent pain-related disability after controlling for characteristic pain intensity. These findings suggest the importance of considering pain catastrophizing as part of a comprehensive assessment of pain in MMTP patients with pain.

The CSP group reported higher levels of catastrophizing than the SP group, and pain status (i.e., SP vs CSP) was an independent predictor of both characteristic pain intensity and pain-related disability. Previous research on this study sample found that the CSP group reported higher levels of pain interference and somatization than the SP group, but the two groups had similar levels of depression, anxiety, and personality disorder criteria [1]. Thus, findings from the current study suggest that 1) catastrophizing may be an important factor that distinguishes CSP and SP groups, and 2) MMTP providers who wish to enhance patients' pain coping skill set might consider assessing and addressing catastrophizing. Notably, cognitive behavioral therapy and physical rehabilitation have demonstrated efficacy in attenuating catastrophizing, which in turn may mediate decreased pain intensity and disability [10,37].

Compton and colleagues, among others, have documented that methadone-maintained patients in comparison to non-addicted individuals exhibit heightened pain sensitivity; our findings suggest that pain catastrophizing may serve as one mechanism that mediates such hyperalgesia [38].

The association between the praying–hoping subscale and characteristic pain intensity approached but did not reach statistical significance. Higher levels of praying–hoping have previously been shown to be associated with higher levels of pain intensity [14,19]. The importance of conducting future studies on praying–hoping in MMTP patients with pain is suggested by the high rates of praying reporting among both the CSP and SP groups. Future studies might benefit from assessing not only the levels of praying–hoping in response to pain, but also the content of the prayers invoked. For example, while some praying may constitute active coping (e.g., “I offer up this pain for the souls in Purgatory”), other types of prayer may be passive (e.g., “Please take away the pain”) or a manifestation of helplessness or catastrophizing (e.g., “Why me, why won't this pain ever go away?”)

The absence of strong relationships between the more active coping strategies and 1) characteristic pain intensity and 2) recent pain-related disability was not unexpected. Previous studies have found inconsistent associations between these strategies and pain intensity and disability [9,39]. In contrast, employment status and pain status were found to be independent predictors of both characteristic pain intensity and pain-related disability. Higher levels of reinterpreting pain sensations among the SP group, in comparison to the CSP group, suggest that future research in this area is merited. For example, one potentially fruitful area for further investigation is assisting patients with co-occurring opioid dependence and chronic pain in MMTPs to gain distance from their pain by imagery-based techniques (e.g., see Pincus and Sheikh [40]).

Several potential limitations of the current study should be acknowledged. The questionnaires relied upon patient self-report, and no independent assessment of pain status, pain intensity, disability, or use of coping strategies was made. The cross-sectional nature of the study design limits statements regarding causality among study variables. Longitudinal studies are needed to assess pain intensity, catastrophizing, pain-related disability, and psychiatric symptoms at different time points to better understand the causal relationship between these variables. We did not collect data regarding the chronology of pain, unemployment, and disability. Future studies might benefit from including this information as it might elucidate whether for some proportion of methadone-maintained patients with pain, a self-perpetuating cycle between pain and disability exists. Not all of the participants who reported pain completed the CSQ-R (108 out of 115 did); the extent to which those who completed the CSQ-R differed from those who did not is unclear. Given that comprehensive pain management services for opioid-dependent MMTP patients with chronic pain will likely require a multidisciplinary approach, future research in this area might benefit from an examination of interventions that are designed to address pain directly (e.g., medications, somatic treatments) in addition to further examination of pain coping strategies.

Despite these limitations, the current study represents an important extension of the research on pain catastrophizing into an MMTP setting, where the prevalence of CSP is particularly high. Consistent with studies of patients with chronic pain who are not opioid dependent, our findings emphasize the importance of assessing and addressing catastrophizing in MMTP patients with pain. Catastrophizing is amenable to cognitive and behavioral interventions, and decreased catastrophizing may help break the cycle of pain followed by catastrophizing, which leads to increased anxiety and distress that further exacerbate pain. Finally, further research in this area might benefit from a more extensive examination of catastrophizing, including social factors and appraisal-related processes [41].

Footnotes

  • 1 We use the terms “opioid dependence” and “opioid dependent” in accordance with Diagnostic Statistical Manual of Mental Disorders, Fourth Edition, Text Revision terminology.

  • This research was supported by funding from the APT Foundation, Inc. and the National Institute on Drug Abuse (K23DA024050; K24 DA0045).

References

View Abstract