Changes in Situation-Specific Pain Catastrophizing Precede Changes in Pain Report During Capsaicin Pain: A Cross-Lagged Panel Analysis Among Healthy, Pain-Free Participants

The Journal of Pain, Vol 11, No 9 (September), 2010: pp 876-884 Available online at www.sciencedirect.com

Changes in Situation-Specific Pain Catastrophizing Precede Changes in Pain Report During Capsaicin Pain: A Cross-Lagged Panel Analysis Among Healthy, Pain-Free Participants Claudia M. Campbell,* Phillip J. Quartana,* Luis F. Buenaver,* Jennifer A. Haythornthwaite,* and Robert R. Edwardsy * Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, Maryland. y Department of Anesthesiology, Perioperative and Pain Medicine, Harvard Medical School, Brigham and Women’s Hospital, Chestnut Hill, Massachusetts.

Abstract: Considerable evidence has linked catastrophizing to pain responses, and recent experimental pain research has suggested that situational catastrophizing, measured during or immediately after laboratory pain procedures, is strongly related to pain ratings of standardized noxious stimuli. However, given that most experimental pain protocols involve ‘‘static’’ assessments of pain ratings and catastrophizing at a single time point, the direction by which these factors may affect each other remains unclear. Does catastrophizing influences one’s subsequent pain responses or do individual differences in the perceived severity of pain lead to differential rates of catastrophizing? Little is known regarding the course of these variables. Using a cross-lagged panel analysis, we evaluated whether changes in situation-specific catastrophizing preceded changes in laboratory-induced pain responses, or vice versa, during tonic capsaicin pain stimulation. Topical application of a 10% capsaicin cream was applied to the dorsal aspect of the nondominant hand of 38 healthy participants. Situation-specific catastrophizing and pain ratings were obtained at Early (0 to 15 minutes), Mid (15 to 30 minutes), and Final (30 to 35 minutes) periods during capsaicin pain. Analyses revealed that Early-to-Mid changes in catastrophizing ratings prospectively accounted for unique variance in subsequent Mid-to-Final changes in pain ratings, whereas Early-to-Mid changes in pain ratings did not account for unique variance in Mid-to-Final changes in catastrophizing ratings. That is, participants who showed the largest initial increases in catastrophizing reported the greatest subsequent increases in pain. Controlling for the reported change in stress did not affect this pattern of results. These findings provide empirical evidence that a situationspecific catastrophizing process might precede and contribute to subsequent increases in pain experience. Limitations of the present study and possible future research directions are discussed. Perspective: The present study adds to a growing literature on prospective associations between catastrophizing and pain. These results provide initial evidence, in healthy individuals, that changes in catastrophizing may precede changes in pain response. ª 2010 by the American Pain Society Key words: Pain, catastrophizing, experimental pain, capsaicin, cross-lagged panel design.

P

ain catastrophizing is well recognized as a prominent risk-factor for inimical acute and chronic pain-related outcomes. Indeed, a self-reported

Received July 7, 2009; Revised December 7, 2009; Accepted December 11, 2009. Supported by NIH grant AT001433 and with resources from F32 NS06362. Address reprint requests to Dr Claudia M. Campbell, Johns Hopkins University School of Medicine, Department of Psychiatry and Behavioral Sciences, 5510 Nathan Shock Drive, Suite 100, Baltimore, MD 21224. E-mail: [email protected] 1526-5900/$36.00 ª 2010 by the American Pain Society doi:10.1016/j.jpain.2009.12.007

876

tendency to catastrophize about pain has been linked to increased pain severity, pain behaviors, and length of hospital stay,38 physical disability,47,50 analgesic use,28 postoperative recovery time,43 suicidal ideation,17 and even the prospective development of pain conditions.9,36,40 More recently, physiological correlates have been linked to catastrophizing, including interleukin6,16 endogenous opioid responsiveness,52 and adrenocortical responses to acute pain.37 Catastrophizing, especially situation-specific catastrophizing during experimental pain testing sessions, has also been associated with greater pain sensitivity to an array of laboratory pain

Campbell et al

The Journal of Pain 18

stimuli, including enhanced temporal summation, cold pressor pain,11,14 and electrically-induced pain responses.20,39 Despite the important role catastrophizing plays in shaping the course of clinical pain, there have been inconsistent findings regarding the construct’s predictive utility with respect to experimental pain responses.22,23,42 Consequently, some researchers have assessed catastrophizing in a situation-specific manner, asking participants to respond to catastrophizing-themed items using a laboratory pain testing experience as the referent. In general, stronger associations between experimental pain responses and situation-specific versus trait catastrophizing have been reported.8,11,14 For example, recent investigations have shown that situational catastrophizing is associated with greater sensitivity to cold pain11,14 as well as other measures of pain thresholds and suprathreshold pain responses in healthy participants.8,39 Three of these published reports also assessed dispositional catastrophizing (ie, questionnaire measures of the degree to which individuals ‘‘usually’’ catastrophize) and found no relationship between experimental pain responses and traditional catastrophizing.8,11,14 Given the post–pain-induction timing of the assessment of situation-specific catastrophizing, it remains unclear whether catastrophizing enhances the experience of pain at a subsequent time point or whether individual differences in pain sensitivity (reflected in variability in pain ratings) drive the degree to which people catastrophize. The current analyses sought to elucidate the direction and nature of this relationship using crosslagged panel analyses31 on data collected during the screening portion of a PET imaging study using a capsaicin pain paradigm. Catastrophizing and pain perception were both measured at multiple points throughout this session, providing a unique opportunity to examine the temporal precedence of catastrophizing in a controlled laboratory setting. The cross-lagged panel analytic approach provides a means to systematically characterize temporal associations between constructs of interest; in this case, situation-specific pain catastrophizing and pain intensity ratings in response to standardized noxious stimuli. Specifically, these analyses allow for the examination of whether changes in one factor (eg, catastrophizing) precede and/or predict subsequent changes in another factor (eg, pain), as well as tests of reverse lagged and concurrent associations. Here, we evaluated whether early changes in situation-specific catastrophizing accounted for subsequent changes in pain report and/or whether early changes in pain report accounted for subsequent changes in situation-specific catastrophizing (Fig 1).

Methods Participants Fifty-four healthy participants were recruited through posted advertisements. Technical difficulties prohibited computerized data collection on 15 individuals, and 1

Δ Pain 1

877

Δ Pain 2

Pain 1

Pain 2

Pain 3

Cat 1

Cat 2

Cat 3

Δ Cat 1

Δ Cat 2

Figure 1. Cross-lagged panel design.

participant did not complete questionnaire data. A total sample size of 38 participants is included in the current analysis. Mean age of this sample was 29.5 years (SD = 10.9), approximately half were female (52.6%) and non-Hispanic white (47.4%). All study-related procedures were approved by the Johns Hopkins Hospital Institutional Review Board. Major eligibility criteria included having no pain condition or medical/psychiatric disorders, alcohol or substance abuse problems, or use of narcotics, antidepressants, anticonvulsants, and muscle relaxants. Verbal and written informed consent was obtained on arrival, after which participants underwent the pain testing procedure described below.

Psychophysical Pain Testing A piece of thick, nonporous dressing with a 6.25 cm2 hole cut into it (used to standardize the area of capsaicin cream application) was applied to the skin on the dorsal nondominant hand. Approximately .35 g of 10% capsaicin cream was applied inside this hole and evenly spread on the skin. The area was then covered by Tegaderm transparent dressing (3M Health Care, St Paul, MN). Because topical capsaicin-induced pain varies strongly as a function of skin temperature, a Peltier device heating element (Medoc US, Minneapolis, MN) was strapped on top of the 6.25 cm2 area with Velcro wrist straps. This device was held at a constant temperature of 38 C for the first 30 minutes of the session. This methodology produces pain that is rated, on average, as moderate in intensity, and that peaks at 15 to 26 minutes after application and plateaus for approximately 1 hour afterward.2,3 Participants provided pain intensity ratings at 30-second intervals on a 0 to 100 computerized visual analogue scale for a total of 35 minutes. During the testing period, participants also practiced playing a series of video games for future sessions (for the parent study) in 5-minute segments (within the 35-minute testing period). Pain ratings were assessed in 5-minute segments between playing video games. Participants played 1 video game during the first time period (between two 5-minute rating blocks) and 2 video games during the second period; thus, 20 ratings were obtained during the Early time period and 5 were obtained during the Mid interval. After 30 minutes had

878

Catastrophizing Precedes Pain

The Journal of Pain

Zero-Order Correlations Between PCS Total Score and Early, Mid, and Final Catastrophizing and Pain Values

Table 1.

HEALTHY (N = 38)

1

2

3

4

5

6

7

1. PCS total score 2. Early Cat 3. Mid Cat 4. Final Cat 5. Early Pain 6. Mid Pain 7. Final Pain

1.0 .26 .24 .26 .09 .17 .13

1.0 .67* .67* .35* .28 .23

1.0 .82* .23 .47* .53*

1.0 .23 .40* .51*

1.0 .47* .35*

1.0 .89*

1.0

Cat, catastrophizing. *P < .05.

elapsed, the thermode temperature was increased to 42  C for an additional 5 minutes. The Final interval comprised these 10 ratings. After completion of the session, the capsaicin cream was removed from the skin.

Psychosocial Questionnaires Pain Catastrophizing Scale Participants completed the Pain Catastrophizing Scale (PCS) just before undergoing pain testing. The measure consists of 13 items rated on a 5-point scale ranging from 0 (not at all) to 4 (all the time).45 Participants are instructed to indicate the degree to which they have specified thoughts and feelings when experiencing pain. The measure assesses 3 dimensions of catastrophizing: rumination, magnification, and helplessness. The PCS has been validated for both clinical and nonclinical samples.34

Situation-Specific Pain Catastrophizing An adaptation of the PCS45 used for a laboratory pain testing environment was used to assess situation-specific catastrophizing. Such questionnaires have been used in a number of studies4,8,11,14,21 and are typically administered just after laboratory pain induction procedures. The Situation-Specific Pain Catastrophizing measure is a 6-item questionnaire with responses ranging from 0 (not at all) to 4 (all the time), with a potential range of scores between 0 and 24. The scale includes questions such as ‘‘I can’t stop thinking about how much it hurts.’’ The scale has been described more fully by Edwards et al,18 and its psychometric properties have recently been investigated.8 In the current study, participants completed the catastrophizing questionnaire at 3 time points while experiencing capsaicin pain: 1) 15 minutes after capsaicin and thermode application, 2) 30 minutes after application, and 3) just before removal of the capsaicin. Participants were instructed to reference the pain they were experiencing in their hand while completing the questionnaire at each time point. Pain ratings were not obtained during the brief period while participants completed the questionnaire.

Pre/Post Stress Questionnaire Participants completed a single question (How much stress do you feel right now?) assessing stress both before

and after the session. These items were rated on a 0 to 10 scale, where 0 = no stress to 10 = extreme stress.

Data Reduction and Analysis Mean pain ratings and summed situation-specific catastrophizing were calculated for each of the 3 time points. Zero-order correlations between the PCS total score and situational catastrophizing and pain ratings at all 3 time points were computed. Three-factor (Early [0 to 15 minutes], Mid [15 to 30 minutes], and Final [30 to 35 minutes] assessments) repeated-measures analysis of variance (ANOVA) was conducted to determine whether changes were observed from Early assessments to Mid and Final assessment epochs for situation-specific catastrophizing and pain ratings. We computed standardized residualized change scores to index Early-to-Mid and Mid-toFinal changes in situation-specific catastrophizing and pain ratings. We opted to use residualized versus simple change scores because of problems of dependence between change and Early values encountered with use of the latter.10 We subsequently examined zero-order correlations between Early-to-Mid and Mid-to-Final changes in catastrophizing and pain. Last, a series of hierarchical regression analyses were conducted to assess whether Early-to-Mid changes in catastrophizing accounted for unique variance in Mid-to-Final change in pain ratings, or vice versa, controlling for synchronous and autocorrelations (ie, potential sources of extraneous variance).

Results As displayed in Table 1, Early, Mid, and Final assessments of catastrophizing were intercorrelated (r = .67 to .82; P < .001), as were pain ratings (r = .35 to .90; P < .05). The PCS total score was not correlated with situational catastrophizing or pain at any time point (consistent with prior studies that have also observed minimal associations between situation-specific and standard catastrophizing).11,14 Early ratings of pain were not associated with catastrophizing at Mid or Final time points, and Early catastrophizing ratings were not significantly associated with pain ratings at Mid or Final time points. Last, Early catastrophizing ratings were positively associated with Early pain ratings, and

Campbell et al Table 2.

The Journal of Pain

879

Early, Mid, and Final Values for Catastrophizing and Pain Ratings

VARIABLE

Catastrophizing Pain

EARLY

MID

FINAL

MEAN

SD

P (EARLY-MID)

MEAN

SD

P (MID-FINAL)

MEAN

SD

P (EARLY-FINAL)

F(2,74)

2.29 8.76

2.88 10.10

.95 <.001*

2.71 19.47

3.35 16.34

.04* <.001*

3.66 30.25

3.91 23.15

.02* <.001*

5.56 32.24

*P < .05.

Mid and Final catastrophizing ratings were both positively related to Mid and Final Pain responses. As shown in Table 2, repeated-measures ANOVA indicated that situational catastrophizing (P = .01) and pain ratings (P < .001) increased significantly over time. Early catastrophizing ratings did not significantly differ from Mid ratings (t = 1.02; P > .10). However, Early catastrophizing ratings significantly differed from Final ratings (t = 2.89; P = .02), and Mid ratings significantly differed from Final ratings (t = 2.58; P = .04). Each pain rating significantly differed from the others. Early pain ratings differed from both Mid and Final ratings (t = 4.50 and 6.08, respectively; P < .001), and Mid ratings differed from Final ratings (t = 5.93; P < .001). No differences emerged between Early-to-Mid and Mid-to-Final changes in catastrophizing or pain responses (P > .01). Correlations among standardized residual change scores are presented in Table 3. Change between Early-to-Mid pain catastrophizing correlated with Early-to-Mid and Mid-to-Final changes in pain.

Cross-Lagged Hierarchical Regression Analyses Hierarchical regressions were performed to determine whether Early-to-Mid pain catastrophizing change scores were significantly and uniquely associated with Mid-to-Final changes in pain ratings, or vice versa. With Mid-to-Final change in pain as the criterion variable, change in pain from Early-to-Mid and change in catastrophizing from Mid-to-Final session accounted for 8% of the variance in the initial step of the regression. In the second step, Early-to-Mid change in catastrophizing accounted for 18% of the variance. These data suggest

Zero-Order Correlation Among PCS Total Score and Standardized Residual Change Scores

Table 3.

VARIABLE

1

2

3

4

5

1. PCS 2. D Cat 1 3. D Cat 2 4. D Pain 1 5. D Pain 2

1.0 .04 .13 .23 .03

1.0 .19 .43* .37*

1.0 .001 .26

1.0 .11

1.0

D Cat 1, Early-to-Mid catastrophizing change; D Cat 2, Mid-to-Final catastrophizing change; D Pain 1, Early-to-Mid pain change; D Pain 2, Mid-to-Final pain change. *P < .05.

that early changes in catastrophizing account for unique variance in later changes in pain (Table 4). With Mid-to-Final change in catastrophizing as the criterion variable, change in catastrophizing from Early-toMid and change in pain from Mid-to-Final accounted for 16% of the variance in the initial step of the regression. In the second step, Early-to-Mid change in pain accounted for only 1% of the variance in Mid-to-Final change in catastrophizing (Table 4).

Post Hoc Analyses Because potential changes in participant stress throughout a session may overlap with catastrophizing to some extent,38 pre, post, and standardized residual change scores for stress from pre to post session were assessed. Pre-session stress was negatively correlated with Early-to-Mid change in catastrophizing (r = .36, P = .03). Post-session stress was associated with the Early catastrophizing ratings (r = .35, P = .02). Change in stress was not associated with changes in catastrophizing or pain at any time point, nor did inclusion of stress as a covariate in the cross-lagged regressions change the pattern of significance. Further, it added only 1% and 1.8% to the total variance accounted for in the initial step of the regressions predicting Mid-to-Final change in catastrophizing and pain, respectively.

Discussion The aim of the present analysis was to examine whether changes in catastrophizing during the pain experience prospectively influence subsequent pain, or vice versa. The present study adds to a growing literature on prospective associations between catastrophizing and pain. Using a cross-lagged panel approach, the change in catastrophizing from Early-to-Mid assessment was associated with (subsequent) changes in Mid-to-Final pain. This finding was observed even when controlling for Mid-toFinal changes in catastrophizing, Early-to-Mid changes in pain, as well as trait catastrophizing and changes in perceived stress. In contrast, changes in pain from Earlyto-Mid were not associated with later changes in Mid-to-Final catastrophizing; however, this lack of significance could in-part be due to the restricted range within catastrophizing scores (which often occurs when studying healthy populations). Further, these results may be influenced or contaminated by the time intervals chosen and the uneven number of pain assessments within each time frame. These results provide preliminary evidence that changes in catastrophizing might precede changes

880

Catastrophizing Precedes Pain

The Journal of Pain

Summary of Hierarchical Regression Analyses: Cross-Lagged Regressions for Catastrophizing Predicting Pain and Pain Predicting Catastrophizing

Table 4. STEP D Pain 2 1

2 STEP D Cat 2 1

2

VARIABLE

R

ADJUSTED R2

R

2

CHANGE

F CHANGE

STANDARDIZED b

P VALUE

.14 .27 .10

.42 .12 .58

D Pain 1 D Cat 2 PCS Total

.3

.006

.09

1.07

D Cat 1

.54

.20

.20

9.17*

VARIABLE

R

ADJUSTED R2

D Cat 1 D Pain 2 PCS Total

.44

.12

D Pain 1

.45

.11

2 R

CHANGE

.53

.005

F CHANGE

STANDARDIZED b

P VALUE

.19

2.65

.36 .40 .16

.04 .02 .32

.01

.48

.12

.49

D Cat 1, Early-to-Mid catastrophizing change; D Cat 2, Mid-to-Final catastrophizing change; D Pain 1, Early-to-Mid pain change; D Pain 2, Mid-to-Final pain change. *P < .05.

in pain response. This statistical approach has been similarly used by Burns et al,5,6 who also found prospective associations of changes in catastrophizing with changes in pain-related variables in the context of multidisciplinary pain treatment. Their results indicated that Early treatment reductions in pain helplessness (a dimension of catastrophizing) predicted Final treatment decreases in pain and interference, and Early treatment reductions in catastrophizing and pain-related anxiety preceded Final treatment improvements in pain severity, but not vice versa.5 They also found that Early treatment changes in catastrophizing and pain helplessness predicted Final treatment pain outcomes, even controlling for depression.6 Conventionally measured catastrophizing purportedly captures a dispositional trait that appears to be relatively stable over time, at least in the absence of intervention.15,30,51 Measurement of catastrophizing is typically assessed before laboratory pain induction procedures, when subjects complete 1 or more questionnaires that ask them to reflect and report on how much they generally catastrophize when in pain; however, recent reports suggest that such dispositional assessment of catastrophizing may not be as relevant to experimentally induced pain, particularly among healthy individuals.8 In part, this finding is likely to be attributable to the ambiguity of the referent and time frame when completing such a questionnaire. Situational catastrophizing is assessed during or immediately after experimental noxious stimulation and refers the participant to the pain experienced during testing. A growing body of literature has noted the strong association between situational catastrophizing and pain ratings in healthy participants as well as chronic pain populations,8,11,18,20,49 and several reports suggest that situational and dispositional measures are only moderately correlated.8,11,14 Our findings suggest that traditionally measured (dispositional) catastrophizing was not associated with situational catastrophizing or pain at any time point during the capsaicin task, nor was

it associated with changes in catastrophizing or pain. Thus it is not surprising that trait catastrophizing did not significantly contribute to the changes in Mid-to-Final pain or catastrophizing. A key issue hampering interpretation of associations between the construct of situational catastrophizing and pain is the timing of the assessment of the situational catastrophizing construct. Specifically, the temporal relationship between pain measurement and situational catastrophizing assessment are very close to each other and sometimes overlap. It is thus difficult to determine whether the relationship between pain report and situational catastrophizing is not in part inflated by the degree of pain experienced. That is, individuals who experience pain more intensely at a given time point might subsequently catastrophize more, producing a significant relationship that is driven by individual differences in pain. Alternatively, we and others have suggested that while the experience of pain surely does influence catastrophizing to a substantial degree (ie, in general, it is unlikely that people catastrophize about pain when they are not experiencing it), the broad individual differences we observe in catastrophizing (even in response to standardized stimulation, or to a pain experience whose rated intensity is identical across individuals) may then causally affect subsequent pain perception. The current study provides preliminary evidence potentially suggesting this directional pattern and may hint at a pathway linking increased catastrophizing with heightened pain response. Consistent with the apparent directionality of our findings (ie, catastrophizing precedes future pain reports), a handful of prospective studies have examined the relationship between catastrophizing and pain in chronic/acute pain populations. For example, Picavet et al36 found that catastrophizing at baseline predicted future chronicity and severity of low back pain. In a randomized controlled trial examining mood and pain in limb loss patients, Hofkamp et al27 found that baseline

Campbell et al catastrophizing was significantly associated with pre- to post-treatment changes in pain; in addition, catastrophizing measured immediately after treatment predicted 3-month post-treatment follow-up pain level. These authors noted that higher catastrophizing appeared to interfere with treatment-associated pain reduction. In an interesting prospective analysis examining catastrophizing during acute dental pain and after the resolution of this pain, Edwards et al15 found that catastrophizing was prospectively associated with diminished thermal pain tolerance over time. They suggested that catastrophizing about pain might promote sensitization to subsequent painful events, even after the resolution of a clinical pain condition. Collectively, these studies hint at a directional influence of catastrophizing and enhanced pain response. In the current study, however, changes in catastrophizing predicted subsequent changes in pain, whereas the converse relationship was not significant. Nevertheless, Early-to-Mid changes in catastrophizing and Mid-toFinal change in pain were significantly predictive of Mid-to-Final changes in catastrophizing. Early changes in catastrophizing predict Final changes in catastrophizing, perhaps suggesting a stable trajectory of situational catastrophizing within person. These results further suggest that, not surprisingly, concurrent changes in catastrophizing and changes in pain are related. Several treatment-related studies have documented this association. For example, Turner et al51 conducted a daily diary study in which pain did indeed predict catastrophizing; however, this relationship was rendered nonsignificant after controlling for Early catastrophizing. In general, numerous studies have shown that although catastrophizing and pain are cross-sectionally related, the association is modest, and large individual differences in catastrophizing are present at any given level of pain intensity.7,13 There appears to be sufficient evidence to hypothesize that those individual differences in catastrophizing have some influence on subsequent pain perception and pain responses, and the present cross-lagged panel analysis hints that changes in catastrophizing are prospectively important as well. That is, over the course of a tonically painful stimulus, those individuals who show the largest initial increases in catastrophizing are those who experience the greatest subsequent increases in pain intensity. As a corollary, those who showed no early increases in catastrophizing had lower pain ratings than those who had early and high levels of catastrophizing. Such findings, in conjunction with Burns et al,6 suggest that targeting catastrophizing at a given time point (either to reduce catastrophizing or prevent its occurrence) is likely to provide later benefits in terms of reduced pain and improved adaptation to pain. The construct of catastrophizing is clearly associated with measures of general distress or negative affect. Pain catastrophizing shares significant variance with broader negative affect constructs, such as depression, anxiety, anxiety sensitivity, worry, and neuroticism,1,25,26,46 though it clearly also has a unique influence on pain outcomes.13,47,48 Although the current

The Journal of Pain

881

investigation was not designed to elucidate the interrelationships between such potentially overlapping phenomena as catastrophizing, stress, and negative mood, it is of interest that pre-post changes in stress was not associated with catastrophizing or pain, nor did these variables affect the cross-lagged relationship between catastrophizing and pain. The mechanisms underlying the influence of catastrophizing on pain remain unclear; however, several hypotheses have been put forth. For example, catastrophizing might be associated with neurophysiological mechanisms that exacerbate the perception of noxious stimuli. Recent reports suggest links between catastrophizing and altered HPA axis responses to acute pain, including pro-inflammatory cytokine IL-6,16 and cortisol,37 though not surprisingly, different patterns of results have emerged between healthy and pain populations.16,29 Functional magnetic resonance imaging studies have also found that catastrophizing is associated with amplification of cortical activation in response to pain.24,41,44 Specifically, higher levels of catastrophizing in pain patients was associated with enhanced pain-related hemodynamic responses in the dorsolateral prefrontal cortex, dorsal anterior cingulate cortex, and medial frontal cortex,24 and when healthy subjects were exposed to mildly painful electrical stimulation, they showed a significant positive relationship between pain catastrophizing scores and pain-related brain responses in the dorsolateral prefrontal cortex, insula, and anterior cingulate cortex,41 brain regions associated with the mediation of emotional and motivational responses to the experience of pain. In addition, our group recently found that situational catastrophizing cognitions measured during the experience of pain were associated with a network of brain regions involving dorsal anterior cingulate cortex, anterior insula, and prefrontal cortex.12 Collectively, these studies suggest that catastrophizing may be associated with amplification of neurophysiological responses to pain. A number of limitations should be considered when interpreting these data. First, consistent with most experimentally induced noxious stimuli assessments, the generalizability of this work to chronic pain patients is unclear as a young, healthy population was recruited. However, use of a healthy population is often an advantage, given the complex factors that often coexist in patients with pain, it provides a ‘‘cleaner’’ method for interrogation. Although many studies provide evidence for the clinical relevance of laboratory pain induction procedures,19 and specifically hyperalgesic responses to topical capsaicin have been associated with clinical pain and tenderness in patients with fibromyalgia and rheumatoid arthritis patients,32,33 further study of the role that catastrophizing might play in modulating pain processes in the presence of clinical pain is certainly warranted. The current findings suggest that catastrophizing about pain may play a role in the perception of future pain intensity. Another limitation could be the restricted range within catastrophizing responses. In cases of restricted range, statistically significant coefficients should most likely be taken

882

The Journal of Pain

Catastrophizing Precedes Pain

seriously, but a nonsignificant coefficient may be weak evidence for the absence of an effect as the magnitude of correlations may be stifled. Thus, the relatively low levels of catastrophizing endorsed by our participants may have made it difficult to detect meaningful relationships between catastrophizing and pain testing variables in the second regression equation. However, it is quite common for healthy participants to endorse low levels of catastrophizing.8,18 Another limitation relates to the use of only a single type of experimental pain stimulus. It is possible that the association between catastrophizing and capsaicin-induced pain responses could vary as a function of the type or intensity of noxious stimulus presented. Thus, assessment of these temporal relationships within and between other experimental pain paradigms should be considered. In addition, the cross-lagged associations reported here followed a planned increase in thermal stimulation. Similar to cross-lagged reports assessing pain treatment over Early to Mid to Final time points, modulation of pain may affect its stability. Another limitation arising from our chosen statistical approach may include the synchronicity of the pain and catastrophizing ratings obtained. While catastrophizing ratings were obtained at specified time intervals, pain ratings were averaged over periods corresponding with those intervals. However, ‘‘retrospective measures,’’ such as collecting ratings over a given time period, are not uncommon in panel studies,31 though future research may wish to address this issue. An additional limitation may be that video games were played during the 35-minute analysis period, for purposes of the parent study. No pain ratings were collected while participants were performing these activities. Another limitation may stem from increasing the temperature after the Mid assessment point. This aspect of our design may have made it more likely to observe a prospective association between catastrophizing and pain rather than vice versa. Experiencing very little pain initially is a function of the capsaicin paradigm and we believe a relative strength of this study. Specifically, even through pain was not at a peak, we achieved notable and apparently

meaningful increases in state catastrophizing. An additional limitation may be that stress was not queried at the same intervals as catastrophizing; it was assessed only pre- and post-pain assessment. As a result, we were unable to determine whether Early-to-Mid stress changes accounted for any portion of the Mid-to-Final pain ratings. In addition, one question related to stress is probably inadequate to gauge emotional distress and a more thorough investigation of this concept should be utilized in future studies. This is an important area for future study, given the conceptual and empirical overlap between catastrophizing, pain, and affect. Another potential limitation may be the relatively brief and uneven time period over which assessment of catastrophizing and pain occurred; however, understanding within-session variability and temporal relationships is an important step toward elucidating these factors over longer time frames and across different populations. Although the repeated assessments are a strength of the current analysis, as this allowed for an evaluation of changes within each construct and how these changes impacted the other, some research would argue for continuous time modeling.35 Nonetheless, these findings suggest that catastrophizing plays an important role in shaping the experience and intensity of pain. These limitations notwithstanding, the current findings are the first of which we are aware that suggest a directional pattern between catastrophizing and pain in the context of a controlled experimental design. The present results provide compelling evidence that changes in catastrophizing might indeed precede changes in pain. Further examination of the role catastrophizing might play in modulating processes in the presence of clinical pain appears needed. Future studies may benefit from the manipulation of the degree of catastrophizing experienced during laboratory testing, using predictive models in a chronic pain population to establish causal pathways, and assessing neurophysiological processes that may underlie the negative impact of catastrophizing on pain perception. Collectively, these findings advance our understanding of the negative effects of catastrophizing on the experience of pain.

References

with acute pain: An examination of the role of endogenous opioid mechanisms. J Behav Med 19:129-142, 1996

1. Affleck G, Tennen H, Urrows S, Higgins P: Neuroticism and the pain-mood relation in rheumatoid arthritis: Insights from a prospective daily study. J Consult Clin Psychol 60: 119-126, 1992

5. Burns JW, Glenn B, Bruehl S, Harden RN, Lofland K: Cognitive factors influence outcome following multidisciplinary chronic pain treatment: A replication and extension of a cross-lagged panel analysis. Behav Res Ther 41:1163-1182, 2003

2. Anderson WS, Sheth RN, Bencherif B, Frost JJ, Campbell JN: Naloxone increases pain induced by topical capsaicin in healthy human volunteers. Pain 99:207-216, 2002

6. Burns JW, Kubilus A, Bruehl S, Harden RN, Lofland K: Do changes in cognitive factors influence outcome following multidisciplinary treatment for chronic pain? A cross-lagged panel analysis. J Consult Clin Psychol 71:81-91, 2003

3. Bencherif B, Fuchs PN, Sheth R, Dannals RF, Campbell JN, Frost JJ: Pain activation of human supraspinal opioid pathways as demonstrated by [11C]-carfentanil and positron emission tomography (PET). Pain 99:589-598, 2002

7. Campbell CM, Edwards RR: Mind-body interactions in pain: The neurophysiology of anxious and catastrophic pain-related thoughts. Transl Res 153:97-101, 2009

4. Bruehl S, Carlson CR, Wilson JF, Norton JA, Colclough G, Brady MJ, Sherman JJ, McCubbin JA: Psychological coping

8. Campbell CM, Kronfli T, Buenaver LF, Haythornthwaite JA, Smith MT, Edwards RR. Situational vs. dispositional measurement of catastrophizing: Associations with pain responses in multiple samples. J Pain11:443-53, 2010

Campbell et al 9. Carroll LJ, Cassidy JD, Cote P: The role of pain coping strategies in prognosis after whiplash injury: Passive coping predicts slowed recovery. Pain 124:18-26, 2006 10. Cohen P, Cohen J, West SG, Aiken LS: Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences, 3rd ed. Mahwah, NJ, Lawrence Erlbaum Associates, Inc, 2001

The Journal of Pain

883

25. Hirsh AT, George S, Riley JLI, Robinson ME: Sex differences and construct redundancy of the Coping Strategies Questionnaire: Catastrophizing subscale. J Pain 6:S60, 2005 26. Hirsh AT, George SZ, Riley JL III, Robinson ME: An evaluation of the measurement of pain catastrophizing by the coping strategies questionnaire. Eur J Pain 11:75-81, 2007

11. Dixon KE, Thorn BE, Ward LC: An evaluation of sex differences in psychological and physiological responses to experimentally-induced pain: A path analytic description. Pain 112:188-196, 2004

27. Hoffkamp SE, Wegener S, MacKenzie EJ, Ephraim P, Ehde DM, Williams RM, Klick B: Changes in catastrophizing predict pain ratings in persons with limb loss. J Pain 9:P52, 2008

12. Edwards RR: Physiological Consequences of Catastrophizing about Pain. J Invest Med 53(3):610, 2008

28. Jacobsen PB, Butler RW: Relation of cognitive coping and catastrophizing to acute pain and analgesic use following breast cancer surgery. J Behav Med 19:17-29, 1996

13. Edwards RR, Bingham CO III, Bathon J, Haythornthwaite JA: Catastrophizing and pain in arthritis, fibromyalgia, and other rheumatic diseases. Arthritis Rheum 55:325-332, 2006 14. Edwards RR, Campbell CM, Fillingim RB: Catastrophizing and experimental pain sensitivity: Only in vivo reports of catastrophic cognitions correlate with pain responses. J Pain 6:338-339, 2005 15. Edwards RR, Fillingim RB, Maixner W, Sigurdsson A, Haythornthwaite J: Catastrophizing predicts changes in thermal pain responses after resolution of acute dental pain. J Pain 5:164-170, 2004 16. Edwards RR, Kronfli T, Haythornthwaite JA, Smith MT, McGuire L, Page GG: Association of catastrophizing with interleukin-6 responses to acute pain. Pain 140:135-144, 2008 17. Edwards RR, Smith MT, Kudel I, Haythornthwaite J: Painrelated catastrophizing as a risk factor for suicidal ideation in chronic pain. Pain 126:272-279, 2006 18. Edwards RR, Smith MT, Stonerock G, Haythornthwaite JA: Pain-related catastrophizing in healthy women is associated with greater temporal summation of and reduced habituation to thermal pain. Clin J Pain 22:730-737, 2006

29. Johansson AC, Gunnarsson LG, Linton SJ, Bergkvist L, Stridsberg M, Nilsson O, Cornefjord M: Pain, disability and coping reflected in the diurnal cortisol variability in patients scheduled for lumbar disc surgery. Eur J Pain 12:633-640, 2008 30. Keefe F, Brown G, Wallston K, Caldwell D: Coping with rheumatoid arthritis pain: Catastrophizing as a maladaptive strategy. Pain 37:51-56, 1989 31. Kenny DA: Cross-lagged panel correlation: A test for spuriousness. Psychol Bull 82:887-903, 1975 32. Morris V, Cruwys S, Kidd B: Increased capsaicin-induced secondary hyperalgesia as a marker of abnormal sensory activity in patients with fibromyalgia. Neurosci Lett 250: 205-207, 1998 33. Morris VH, Cruwys SC, Kidd BL: Characterisation of capsaicin-induced mechanical hyperalgesia as a marker for altered nociceptive processing in patients with rheumatoid arthritis. Pain 71:179-186, 1997 34. Osman A, Barrios FX, Gutierrez PM, Kopper BA, Merrifield T, Grittmann L: The Pain Catastrophizing Scale: Further psychometric evaluation with adult samples. J Behav Med 23:351-365, 2000

19. Fillingim RB, Lautenbacher S: The importance of quantitative sensory testing in the clinical setting, in Lautenbacher, S. and Fillingim, R.B. (eds): Pathophysiology of Pain Perception, Plenum Series in Rehabilitation and Health. New York, NY, Kluwer Academic/Plenum Publishers, 2004, pp 215-225

35. Oud JHL: Continuous time modeling of reciprocal relationships in the cross-lagged panel design, in Boker S.M. and Wenger M.J. (eds): Data Analytic Techniques for Dynamical Systems, Notre Dame Series on Quantitative Methodology. Mahwah, NJ, Lawrence Erlbaum Associates, Inc., 2007, pp 87-129

20. France CR, France JL, Al’absi M, Ring C, McIntyre D: Catastrophizing is related to pain ratings, but not nociceptive flexion reflex threshold. Pain 99:459-463, 2002

36. Picavet HS, Vlaeyen JW, Schouten JS: Pain catastrophizing and kinesiophobia: Predictors of chronic low back pain. Am J Epidemiol 156:1028-1034, 2002

21. Geisser ME, Robinson ME, Pickren WE: Differences in cognitive coping strategies among pain-sensitive and paintolerant individuals on the cold pressor test. Behav Ther 23:31-42, 1992

37. Quartana PJ, Buenaver LB, Edwards RR, Klick B, Haythornthwaite JA, Smith MT: Pain catastrophizing and salivary cortisol responses to laboratory pain testing in temporomandibular disorder and healthy participants. J Pain 11:186-94, 2010

22. George SZ, Dannecker EA, Robinson ME: Fear of pain, not pain catastrophizing, predicts acute pain intensity, but neither factor predicts tolerance or blood pressure reactivity: An experimental investigation in pain-free individuals. Eur J Pain 10:457-465, 2006 23. George SZ, Hirsh AT: Psychologic influence on experimental pain sensitivity and clinical pain intensity for patients with shoulder pain. J Pain 10:293-299, 2009 24. Gracely RH, Geisser ME, Giesecke T, Grant MA, Petzke F, Williams DA, Clauw DJ: Pain catastrophizing and neural responses to pain among persons with fibromyalgia. Brain 127:835-843, 2004

38. Quartana PJ, Campbell CM, Edwards RR: Pain catastrophizing: A critical review. Exp Rev Neurotherapeut 9: 745-758, 2009 39. Rhudy JL, Maynard LJ, Russell JL: Does in vivo catastrophizing engage descending modulation of spinal nociception? J Pain 8:325-333, 2007 40. Richardson C, Glenn S, Horgan M, Nurmikko T: A prospective study of factors associated with the presence of phantom limb pain six months after major lower limb amputation in patients with peripheral vascular disease. J Pain 8: 793-801, 2007

884

The Journal of Pain

41. Seminowicz DA, Davis KD: Cortical responses to pain in healthy individuals depends on pain catastrophizing. Pain 120:297-306, 2006 42. Severeijns R, Van den Hout MA, Vlaeyen JW: The causal status of pain catastrophizing: An experimental test with healthy participants. Eur J Pain 9:257-265, 2005 43. Stephens MA, Druley JA, Zautra AJ: Older adults’ recovery from surgery for osteoarthritis of the knee: Psychosocial resources and constraints as predictors of outcomes. Health Psychol 21:377-383, 2002 44. Strigo IA, Simmons AN, Matthews SC, Craig AD, Paulus MP: Increased affective bias revealed using experimental graded heat stimuli in young depressed adults: Evidence of ‘‘emotional allodynia’’. Psychosom Med 70: 338-344, 2008 45. Sullivan MJ, Bishop S, Pivik J: The Pain Catastrophizing Scale: Development and validation. Psychol Assess 7: 524-532, 1995 46. Sullivan MJ, D’Eon JL: Relation between catastrophizing and depression in chronic pain patients. J Abnorm Psychol 99:260-263, 1990

Catastrophizing Precedes Pain 47. Sullivan MJ, Thorn B, Haythornthwaite JA, Keefe F, Martin M, Bradley LA, Lefebvre JC: Theoretical perspectives on the relation between catastrophizing and pain. Clin J Pain 17:52-64, 2001 48. Sullivan MJ, Thorn B, Rodgers W, Ward LC: Path model of psychological antecedents to pain experience: Experimental and clinical findings. Clin J Pain 20: 164-173, 2004 49. Thorn BE, Clements KL, Ward LC, Dixon KE, Kersh BC, Boothby JL, Chaplin WF: Personality factors in the explanation of sex differences in pain catastrophizing and response to experimental pain. Clin J Pain 20:275-282, 2004 50. Turner JA, Jensen MP, Warms CA, Cardenas DD: Catastrophizing is associated with pain intensity, psychological distress, and pain-related disability among individuals with chronic pain after spinal cord injury. Pain 98:127-134, 2002 51. Turner JA, Mancl L, Aaron LA: Pain-related catastrophizing: A daily process study. Pain 110:103-111, 2004 52. Weissman-Fogel I, Sprecher E, Pud D: Effects of catastrophizing on pain perception and pain modulation. Exp Brain Res 186:79-85, 2007