PAIN Publish Ahead of Print DOI: 10.1097/j.pain.0000000000000519

Original article Experimental manipulations of pain catastrophizing influence pain levels in chronic pain patients and healthy volunteers

D

Heidi Kjøgxa,b, Helge Kaschb,c, Robert Zachariaea,d, Peter Svenssone.f, Troels S. Jensenb, Lene Vasea,b*, a

TE

Department of Psychology and Behavioural Sciences, Aarhus University, Aarhus, Denmark Danish Pain Research Center, Aarhus University Hospital, Aarhus, Denmark c Research Department, Spinal Cord Injury Center of Western Denmark, Department of Neurology, Regional Hospital of Viborg, Denmark d Department of Oncology, Aarhus University Hospital, Aarhus, Denmark e Section of Orofacial Pain and Jaw Function, School of Dentistry, Aarhus University, Aarhus, Denmark f Department of Dental Medicine, Karolinska Institutet, Huddinge, Sweden

C EP

b

Summary: Hypnotic suggestions of increased and decreased pain catastrophizing influenced the subsequent pain ratings in chronic pain patients and healthy volunteers

C

Keywords: Pain Catastrophizing, Hypnotic suggestions, Pain Catastrophizing Scale

A

Number of pages: 31 Number of tables: 4 Number of figures: 4

*Correspondence Lene Vase Department of Psychology and Behavioural Sciences Aarhus University Bartholins Allé 9, Building 1350, DK-8000 Aarhus C, Denmark Tel.: +45 87165828. E-mail: [email protected]

1 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

The authors report no conflict of interest.

Abstract Pain catastrophizing (PC) has been related to pain levels in both patients experiencing acute or chronic pain and in healthy volunteers exposed to experimental pain. Still, it is unclear whether high levels of pain catastrophizing lead to high levels of pain or vice versa. We therefore tested if levels

D

of pain catastrophizing could be increased and decreased in the same participant via hypnotic suggestions and whether the altered level of situation-specific pain catastrophizing was related to

TE

increased and decreased pain levels, respectively. Using the spontaneous pain of twenty-two chronic tension-type headache patients and experimentally induced pain in twenty-two healthy volunteers, participants were tested in three randomized sessions where they received three types of

C EP

hypnotic suggestions: Negative (based on the 13 items in the Pain Catastrophizing Scale (PCS)), Positive (coping-oriented reversion of the PCS) and Neutral (neutral sentence) hypnotic suggestions. The hypnotic suggestions significantly increased and decreased situation-specific PC in both patients and healthy volunteers (p < 0.001). Also, the levels of pain intensity and pain unpleasantness were significantly altered in both patients and healthy volunteers (p < 0.001).

C

Furthermore, regression analyses showed that changes in pain catastrophizing predicted changes in pain in patients (R2= 0.204- 0.304; p < 0.045) and in healthy volunteers (R2= 0.328-0.252; p
2 on the NRS, neurological diseases, mental illness, substance abuse (drugs, alcohol or medicine) or lack of ability to cooperate during the experiment. One hundred and eleven (111) healthy volunteers responded to the flyer, 23 fulfilled the inclusion criteria and one subsequently withdrew (See Appendix 1).

5 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

All participants were asked to refrain from pain medication 24 hours prior to each session and none of the participants reported taking pain medication prior to any of the session. All participants gave signed informed consent and the patients received no payment, whereas the healthy volunteers received a small gift (wine) after each of the three sessions. The local Ethics Committee for the Central Denmark Region (#1107232312) approved the study, which adhered to the principles set

D

out by the Helsinki Declaration. All participants were enrolled between April 2013 and April 2014 and tested at the Headache Clinic, Department of Neurology, Aarhus University Hospital, Denmark

TE

by HK. Patients were blinded to the purpose of the experiment. A research assistant assigned participants to the interventions based on computer-randomization and sealed envelopes.

C EP

2.2. Conditions and hypnotic manipulations

All participants were subjected to three experimental conditions, thereby making it a withinparticipant design. In each condition, one of the following types of hypnotic suggestions, were given in a randomized order: Negative Condition, Positive Condition or Neutral Condition. In the Negative Condition, the hypnotic suggestions were based directly on the 13 items of the PCS,

C

and rephrased into hypnotic suggestions directed at the participant. For example, PCS item 3: “It’s terrible and I think it’s never going to get any better” was rephrased in order to become more

A

personal: “You think that it’s terrible – it will never get better”. In the Positive Condition, the hypnotic suggestions were also based on the PCS, but with the content reversed from negative to positive and again rephrased into a hypnotic suggestion directed at the participant. For example, PCS item 3: “It’s terrible and I think it’s never going to get any better” was reversed to a positive and personal form: “You think that actually it’s no big deal – it will get better very soon.”

6 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

In the Neutral Condition, the hypnotic suggestion was the sentence: You will continue to find yourself in a neutral state. See appendix 2 for specification of all suggestions. In the PCS, PC and pain are closely linked so it can be discussed whether the hypnotic suggestions target PC only. Yet, previous experimental studies have manipulated the level of PC using items from the PCS.32 Also, when directly comparing the hypnotic suggestions aimed at manipulating PC

D

(i.e. “you think that it is terrible – it will never get better”), with hypnotic suggestions aimed at manipulating the pain experience itself (i.e”imagine how that specific area has blocked every

TE

sensation like all nerves in that area were cut like a wire and no longer able to pass any sensation on”),1 the difference between targeting cognitive and emotional aspects of PC as opposed to

C EP

targeting the sensory discriminative aspect of pain is specified.

2.3 Predictors: pain catastrophizing measures 2.3.1 Dispositional pain catastrophizing

Dispositional PC was assessed using the validated Danish version of the PCS.14,28 In this questionnaire, participants are informed that “We are interested in the types of thoughts and feelings

C

that you have when you are in pain” and participants are then instructed to rate specific thoughts and feelings when they experienced pain (in general) on a 5-point Likert rating scale. The PCS

A

consists of 13 items covering three subscales: rumination (items 8-11), magnification (items 6, 7, 13) and helplessness (items 1-5, 12).

2.3.2 Situation-specific pain catastrophizing (PCS-3) We were also interested in obtaining situation-specific PCS scores during the hypnotic trance. As the painful stimuli had to be maintained at a stable level throughout the hypnotic trance, we chose for ethical and practical reasons to reduce the 13 items in the PCS to three items (PCS-3)

7 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

representing each of the three subscales of the PCS (rumination, magnification and helplessness). We carefully selected the items that showed the highest factor loadings on each of the three subscales of the Danish version of the PCS.14 The three items were: Number 5: “I feel that I can’t stand it anymore” (helplessness), number 6: “I become afraid that the pain will get worse” (magnification) and number 10: “I keep thinking about how much it hurts” (rumination). Prior to

D

answering the PCS-3 the participants were told: “We are interested in the types of thoughts and feeling that you have during your current pain”. In this way the items resemble a situation-specific

TE

measure of PCS.12,32 The three items were read out aloud and rated on a 5-point Likert scale (0-4), by having the participant raise their index finger when the number that corresponded to their experience was read out. The instructions were: Simply let your body answer by letting your index

C EP

finger lift itself at the number that describes your experience the best. PCS-3 was the predefined primary predictor.

2.4 Outcome measures: Pain 2.4.1 Pain model

C

For chronic pain patients, the patients’ own spontaneous chronic tension-type headache pain level was measured throughout the study. In the healthy volunteers an experimental pain that mimics

A

some aspects of tension-type headache was induced and measured throughout the study. Pain was induced by giving an intramuscular infusion of hypertonic saline (HS) (5%) into the right side of the neck in level with and 2 cm lateral to the spinous process of the third cervical vertebra.30 A disposable 0.7 x 19 mm stainless steel needle (BD Neoflon, RF391350) was placed in the splenius capitis muscle at the specific location in a depth of approximately 19 mm; the needle was placed in a slight cranial direction to ease flow and fixed with surgical tape. The needle was connected to an infusion pump (Alaris, Asena GH, MKIII) via a tube (BD Connecta, RF397590) and a 10 ml plastic

8 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

syringe (BD Plasticpak, RF 302188). The infusion paradigm was 0.5 ml HS infused over 20 sec as a bolus followed by a steady infusion rated of 7 ml/h for approximately 10 minutes.24,30 During pilot testing, care was taken to match the level of experimental pain in healthy volunteers with the level of spontaneous chronic tension-type headache pain in patients (see appendix 4; in the experiment

D

the pain levels were matched).

2.4.2. Pain measurements

TE

Both the sensory (intensity) and affective (unpleasantness) dimensions of pain were measured and the distinction between the two was explained according to standardized statements.21 Pain intensity and unpleasantness were measured using two separate numerical rating scales (NRS) ranging from

C EP

0 (‘no pain sensation’/’not at all unpleasant’) to 10 (‘the most intense pain sensation imaginable’/’the most unpleasant imaginable’).21

2.5 Procedure Prior to the study, the PCS, TAS, and the Diagnostic Criteria for Temporomandibular Disorders (DC/TMD)/Headache Questionnaire23 (for further description see below) were mailed out to

C

potential participants. If the participants were eligible for inclusion, they were invited to participate in the study at the Headache Clinic. The study took place at the Headache Clinic, Aarhus University

A

Hospital, Aarhus, Denmark. The three sessions were held approximately one week apart, in order to avoid carry-over effects. A computer randomized the sessions, and prior to the first session each participant was invited to draw an envelope with a randomization code for the order of sessions. Participants were blinded to the order of the sessions. Each session proceeded as follows: The participants were greeted by the experimenter wearing a doctor’s white coat and were escorted to the examination room and introduced to the study. In the first session the experimenter took time to establish a good relationship and demographic data were obtained along with a detailed history of 9

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

pain. Participants were then introduced to the pain scales. In each session participants completed the dispositional PCS and NRS for pain intensity and unpleasantness (see above). Subsequently, the patients were invited to lie down on a hospital bed for the hypnotic session and in the healthy volunteers an intramuscular catheter was inserted into the neck muscle before they were invited to

D

lie down.

Hypnotic induction: The participants were asked to rub their hands together vigorously and when

TE

experiencing warmth between their hands they were told to imagine they were holding a ball between their hands. This was done to facilitate the induction by combining two sensory modalities, visual and auditory.18 After the participants had acknowledged that they imagined a ball between

C EP

their hands, further relaxation through deep breaths and body scans were completed. When relaxed, a countdown, including a pattern interrupting confusion technique (i.e. counting 15, 14, 15, 14, 13) was used to focus attention and deepen the hypnotic trance (see appendix 3 for hypnotic induction). For healthy volunteers the bolus was given and the infusion of HS was started.

Hypnotic state and manipulations: Once the hypnotic state was induced, but prior to the actual

C

manipulation with hypnotic suggestions, participants were asked to answer three items from the

A

situation-specific PCS (see above) and the NRS for pain intensity and unpleasantness, by raising their index finger as described above. Then the manipulation of pain catastrophizing with hypnotic suggestions was initiated. In the Negative Condition or Positive Condition, all 13 items from the PCS or the reversed PCS (see above), were read aloud twice to the participants, for a duration of approximately 10 minutes. In the Neutral Condition the neutral suggestion (see above) was repeated for approximately 10 minutes. Following the manipulation, the patients were again asked to answer PCS-3 and the NRS for pain intensity and unpleasantness by raising their finger at the relevant

10 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

number between 0 and 10 that was read out loud to them. Importantly, measures of situationspecific PCS were always obtained prior to NRS pain ratings. After this the healthy volunteers had the saline infusion stopped. Termination of hypnosis: Hypnosis was terminated by counting upwards and instructing the participants that the hypnosis had ended and that they would feel refreshed and alert. The

D

participants were then asked to close their eyes again and relaxing non-hypnotic suggestions were given to ensure that no residual pain or unpleasantness would remain. Once the participants were

TE

de-hypnotized and had received the relaxing non-hypnotic suggestions, the intramuscular catheter was removed from the healthy volunteers, and participants were escorted back to the experimenter’s desk and were asked to complete the dispositional PCS and NRS for pain intensity and

C EP

unpleasantness. To ensure that an acceptable level of hypnotic trance had been reached, the participants were asked to score their own level of hypnotic trance on a scale of 0 (“no trance”) to 10 (“deep hypnotic state”). For an overview of the experimental procedure, see Figure 1.

C

(Insert Figure 1)

2.6 Additional Questionnaires

A

2.6.1 Diagnostic Criteria for Temporomandibular Disorders (DC/TMD)/Headache Questionnaire The Headache Questionnaire is a 12-item subscale of the DC/TMD23 focusing specifically on headache. The questionnaire was used to include patients and exclude healthy volunteers with significant neck, jaw and head pain. The twelve questions related to severity, duration, frequency and daily impact of headache.

2.6.2 Tellegen Absorption Scale (TAS)

11 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

The Tellegen Absorption Scale (TAS) was used to measure absorption, which has been found to correlate positively with hypnotizabillity.9,37 It is a 34-item questionnaire derived from the Multidimensional Personality Questionnaire50 and possible answers are either “yes” or “no”.

2.6.3 Harvard Group Scale of Hypnotic Susceptibility, Form A (HGSHS)

D

The Harvard Group Scale of Hypnotic Susceptibility, Form A27 has been developed to permit larger groups of participants to be tested for hypnotic susceptibility. The manual has instructions for

TE

hypnotic induction and for measuring hypnotic capacity on a 12-point scale. The HGSHS has been validated for use in Danish populations, where the total scale reliability has been found acceptable (0.70).36

C EP

The HGSHS was used to test the hypnotic susceptibility of the healthy volunteers before inclusion. For practical reasons it was not possible to test the patients using the HGSHS. Both patients reported and observer reported evaluation of hypnotic trance at the end of the study indicated that all participants had reached an acceptable level of hypnotic trance during the experiment.

2.7. Statistical analysis

C

All analyses were conducted using SPSS 21. Two-tailed p-values less than 0.05 were considered statistically significant. The data were tested for normality using histograms and Q-Q plots. Results

A

were normally distributed and are expressed in mean values ± standard deviations. The primary aim was to test if PC could be experimentally manipulated in positive and/or negative directions via hypnotic suggestions. The primary predictor was the situation-specific PCS scores obtained during hypnosis. Pre- and post-manipulation scores for situation-specific PCS were calculated for each of the three conditions: Negative Condition, Positive Condition and Neutral Condition. The difference scores (post-pre manipulation) in these three conditions were compared

12 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

using repeated-measures analyses of variance (ANOVA). Effect sizes were calculated as Cohen’s d. The secondary aim was to test if a change in PC corresponded to a change in pain measures. In order to be able to compare the situation-specific PCS (post-pre) difference scores, with NRS pain (post-pre) difference scores, NRS pain difference scores were calculated for each of the three

D

conditions: Negative Condition, Positive Condition and Neutral Condition. To assess if changes in pain catastrophizing would predict changes in pain levels, regression analyses were conducted with

TE

situation-specific PCS difference scores (post-pre) as the independent variable and NRS difference score (post-pre) for pain intensity and unpleasantness, respectively, as the dependent variables. Separate regressions were conducted for each variable.

C EP

In all analyses, patients and healthy volunteers are placed side-by-side for descriptive purposes only (cf. introduction). The only exception being post-hoc test in which patients and healthy volunteers were compared with respect to change in situation-specific PCS difference scores (post-pre) and NRS difference scores (post-pre) from the Neutral Condition to the Negative Condition as well as from the Neutral Condition to the Positive Condition. To do so changes in situation-specific PCS difference score (post-pre) from the Neutral Condition to the Negative Condition were calculated

C

for patients and healthy volunteers, respectively, and compared using Student’s t-test. Similarly, the

A

changes in situation-specific PCS difference score (post-pre) from the Neutral Condition to the Positive Condition were calculated and compared for the two groups. The NRS difference scores (post-pre) were calculated and compared in a similar manner.

3.Results 3.1. Participant characteristics A total of 22 pain patients, 18 women and 4 men, with an average age of M = 30.0/SD = 8.8 years, as well as 22 healthy volunteers, 13 women and 9 men, with an average age of M = 23.3 / SD = 4.9 13 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

years participated in this study. Demographic data, dispositional pain catastrophizing (PCS), pain intensity (NRS) and pain unpleasantness (NRS) ratings at the time of inclusion are presented in Table 1.

D

(Insert Table 1 here)

3.2 Pain catastrophizing

TE

3.2.1. Patients

The ANOVA comparisons for the situation-specific PCS difference score (post-pre) ratings in Negative Condition, Positive Condition and Neutral Conditions were significant (F (2,59) = 62.07,

C EP

p < .001). Post hoc tests with Bonferroni corrections and visual inspection of Fig. 2 showed that the mean situation-specific PCS difference score (post-pre) was significantly higher in the Negative Condition compared to the Neutral Condition (p < .001), indicating that the patients catastrophized more in the Negative Condition as compared to the Neutral Condition, corresponding to an effect size of d = 1.41. In the Positive Condition, the mean situation-specific PCS difference score (postpre) was significantly lower compared to the Neutral Condition (p < .001), indicating that the

C

patients catastrophized less in the Positive Condition as compared to the Neutral Condition,

A

corresponding to an effect size of d= - 1.44.

3.2.2. Healthy volunteers The ANOVA comparisons for the PCS-3 difference score (post-pre) ratings in Negative Condition, Condition and Neutral Conditions were significant (F (2,63) = 70.73, p < .001). Post hoc tests with

Bonferroni corrections and visual inspection of Fig. 2 showed that the mean situation-specific PCS difference score (post-pre) was significantly higher in the Negative Condition compared to the Neutral Condition (p < .001), indicating healthy volunteers catastrophized more in the Negative

14 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

Condition compared to the Neutral Condition, with the difference corresponding to an effect size of d = 1.11. In the Positive Condition, the mean situation-specific PCS difference score (post-pre) was significantly lower compared to the Neutral Condition (p < .001), indicating that healthy volunteers catastrophized less in the Positive Condition compared to the Neutral Condition, with the difference corresponding to an effect size of d = -1.54.

D

(Insert Fig. 2 here)

TE

3.3 Pain ratings

3.3.1. Patients

The ANOVA comparisons for the NRS difference score (post-pre) in Negative Condition, Positive

C EP

Condition and Neutral Conditions were significant for both pain intensity (F (2,59) = 22.80, p < .001) and pain unpleasantness (F (2,59) = 33.27, p < .001). Post hoc tests with Bonferroni corrections and visual inspection of Fig. 3 and 4 showed that mean pain intensity and unpleasantness levels were significantly higher in the Negative Condition compared to the Neutral Condition (pain intensity p = .011, pain unpleasantness p = .013), indicating that patients

C

experienced higher pain levels in the Negative Condition compared to the Neutral Condition, with the difference corresponding to an effect size of d = .62 for pain intensity and d = 0.71 for pain

A

unpleasantness. Mean pain intensity and unpleasantness levels were significantly lower in the Positive Condition compared to the Neutral Condition (pain intensity p < .001, pain unpleasantness p < .001), indicating that patients experienced lower pain levels in the Positive Condition compared

to the Neutral Condition, with the difference corresponding to an effect size of d = -.83 for pain intensity and d = -1.15 for pain unpleasantness.

3.3.2. Healthy volunteers

15 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

The ANOVA comparisons for the NRS difference score (post-pre) in Negative Condition, Positive Condition and Neutral Condition were significant for pain intensity (F (2,63) = 54.387, p < .001) and unpleasantness (F (2,63) = 33.016, p < .001). Post hoc tests with Bonferroni corrections and visual inspection of Fig. 3 and 4 showed that mean pain intensity and unpleasantness levels were significantly higher in the Negative Condition compared to the Neutral Condition (pain intensity p

D

< .001, pain unpleasantness p =.013), indicating that healthy volunteers experienced higher pain levels in the Negative Condition compared to the Neutral Condition, with the difference

TE

corresponding to an effect size of d = .77 for pain intensity and d = .67 for pain unpleasantness. Pain intensity and unpleasantness levels were significantly lower in the Positive Condition compared to the Neutral Condition (pain intensity p < .001, pain unpleasantness p < .001),

C EP

indicating that patients experienced lower pain levels in the Positive Condition compared to the Neutral Condition, with the difference corresponding to an effect size of d = - 1.21 for pain intensity and d = -1.17 for pain unpleasantness.

(Insert Fig. 3 and 4 here, see also appendix 4 for actual pain ratings) 3.4 Pain catastrophizing and pain

C

3.4.1. Patients

A

Regression analyses showed that changes in pain catastrophizing predicted changes in pain in patients. In the Negative Condition, the situation-specific PCS difference score (post-pre) predicted the NRS difference score (post-pre) for pain intensity (B = .327, R2 = .204, p = .045) and for pain unpleasantness (B = .377, R2 = .425, p = .002). Similar, in the Positive Condition, the situationspecific PCS difference score (post-pre), predicted the NRS difference score (post-pre) for pain intensity (B = .408, R2 = .304, p = .012) and for pain unpleasantness (B = .435, R2 = .302, p = .012)

16 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

(see Table 2). Thus, increased and decreased levels of PC were related to increased and decreased levels of pain intensity and unpleasantness, in pain patients. In the Neutral Condition, the situation-specific PCS difference score (post-pre) did not predict the NRS difference score (post-pre) for pain intensity (B = .185, R2 = .36, p = .398) or unpleasantness

D

(B = .256, R2 = .112, p = .128) (see Table 2).

3.4.2. Healthy volunteers

TE

Regression analyses showed that changes in pain catastrophizing predicted changes in pain in healthy volunteers. In the Negative Condition, the situation-specific PCS difference score (post-pre) predicted the NRS difference score (post-pre) for pain intensity (B = .430, R2 = .328, p = .005) and for pain unpleasantness (B = .617, R2 = .342, p = .004). In the Positive Condition, the situation-

C EP

specific PCS difference score (post-pre) predicted the NRS difference score (post-pre) for pain intensity (B = .369, R2 = .251, p = .018) and showed a trend towards significance prediction for pain unpleasantness (B = .363, R2 = .174, p = .053) (see Table 2). Thus, increased and decreased levels of PC were related to increased and decreased levels of pain intensity and pain unpleasantness, in healthy volunteers.

C

In the Neutral Condition, the situation-specific PCS difference score (post-pre) predicted the NRS difference score (post-pre) for pain intensity (B = .699, R2 = .450, p = .001) and for pain

A

unpleasantness (B = .777, R2 = .505, p < .000) (see Table 2). Due to the small changes in the neutral group, a post-hoc comparison of the regression analysis in patients and healthy volunteers was made (see Appendix 3). The results of this comparison showed that there was no significant difference between the two correlations for pain intensity (p = ns), but there was for pain unpleasantness (p = (Z = 2.21), thereby indicating that the relationship between situation-specific PCS difference score (post-pre) and NRS difference score (pain intensity and unpleasantness) may or may not be different in the two groups in the Neutral Condition. The potential differences

17 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

observed between patients and healthy volunteers in the Neutral Condition may be due to the larger variability in the patients pain score as they experience a higher degree of heterogeneity in their spontaneous pain levels than healthy volunteers do in response to the standardized experimental pain stimulus.

D

(Insert Table 2 here)

TE

3.5 Increase and decrease of PCS-3, pain intensity and unpleasantness

In a post-hoc analysis, it was analyzed if there were difference in chronic pain patients and healthy volunteers in terms of the changes in PC and pain levels (see Table 3). The situation-specific PCS

C EP

difference score (pre-post) was compared in the Neutral Condition versus the Negative Condition and also in the Neutral Condition versus the Positive Condition. Patients experienced a significantly larger increase in situation-specific PCS levels as compared to healthy volunteers in the Neutral Condition versus the Negative Condition (p = .016). In the Neutral Condition versus Positive Condition, patients tended to experience a lower improvement in situation-specific PCS scores compared to healthy volunteers, but this difference was not significant (p = .272).

C

There were no significant difference between NRS difference score (pre-post) in the Neutral

A

Condition versus the Negative Condition (pain intensity p = .916, pain unpleasantness p = .690) or in the Neutral Condition versus the Positive Condition (pain intensity p = .260, pain unpleasantness

p = .677), thereby indicating that there was no significant differences in the extent to which pain

levels could be changed in patients and healthy volunteers.

(Insert Table 3 here)

3.6 PCS-dispositional 18 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

Unrelated to the hypnotic manipulations, PCS-dis was measured before and after hypnosis. No significant changes were seen in the dispositional PCS scores before and after hypnosis in patients and healthy volunteers (see Table 4), thereby indicating that the changes observed in the situationspecific PCS difference score (post-pre) during hypnosis were not a general change in the level of

D

PC, but a specific effect of the manipulations.

TE

(Insert Table 4 here)

4. Discussion

This is the first study to successfully manipulate pain catastrophizing in a positive and negative

C EP

direction and to show that these manipulations influence pain levels in the same participant. This is shown for both chronic pain patients and healthy volunteers.

4.1 Pain catastrophizing and pain levels

The study shows that it is possible to both increase and decrease the levels of situation-specific PC in the same participant via negative and positive verbal suggestions reinforced by hypnosis.

C

Importantly, both increased and decreased levels of situation-specific PC were related to increased and decreased pain intensity and unpleasantness levels, respectively. The finding that PC can be

A

increased is in line with the study by Severeijns et al.26 where the level of PC was increased in

healthy volunteers using verbal suggestions. Yet, in the present study, the verbal suggestions were directly based on the PCS and reinforced via hypnosis and the increase in the level of PC was measured on situation-specific PCS. Importantly, in the present study, the changes in the level of PC were related to an increase in the level of pain and to our knowledge this is the first study to show that when the level of PC is increased, the level of pain is also increased. In the present study,

19 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

the participants had moderate levels of PC and it would be interesting to further test if the participants with initially low level of PC would also experience an increase in PC and pain levels. The finding that situation-specific PC can be decreased and that this decrease is associated with the level of pain is consistent with the findings by Terry and colleagues.36 Yet, in the present study, the intervention (i.e. hypnotic suggestion for decrease of pain) was compared to a control condition and the effects were

D

with neutral suggestions, which minimizes the risk of attention bias,

demonstrated on situation-specific PC level and pain levels obtained during experienced pain

TE

(either painful stimuli or spontaneous ongoing pain) which minimizes the risk of recall bias, thereby further corroborating the findings by Terry and colleagues .

Notably, the changes in PC and pain levels were for the first time observed in the same participant,

C EP

thereby providing strong evidence for the relationship between PC and pain as it shows that when the PC levels are altered in either direction, pain levels are also altered. In line with previous studies PC accounted for between 17 and 42 percent of the variance in pain intensity and pain unpleasantness findings5,8,26,27,34(see Table 2).

4.2. Similarities and differences between patients and healthy volunteers

C

The present study also revealed for the first time that experimental manipulations of PC is not only

A

associated with healthy volunteers experience of experimental pain but also with chronic pain patients experience of spontaneous ongoing pain. Descriptively it was observed that the patterns of change in PC and pain were similar in the two populations. Post-hoc, statistical comparisons revealed that the levels of situation-specific PC in chronic pain patients increased more compared to healthy volunteers

but except from this finding, there was no significant difference between

changes in situation-specific PC and changes in pain levels between patients and healthy volunteers. Thus, these findings provide a first tentative answer to the questions raised by Severijn and

20 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

colleagues,26 whether it is equally simple to manipulate levels of PC in patients and in healthy volunteers. It appears to be equally simple to manipulate the situation-specific PC level, but chronic pain patients may be more prone to increased PC levels. Also, pain levels appear to be altered as a result of the manipulations, across both populations. Studies which are powered and precisely matched to directly compare the two populations are, however, needed to further test this.

D

Importantly, in the present study the effect sizes for the change in situation-specific PC was large (Cohen’s d =1.11-1.54) and the change in relation to pain was moderate to large (Cohen’s d = 0.62-

TE

0.77).5 The changes in chronic pain patients’ spontaneous pain levels corresponded to an effect size of 0.83. It has previously been suggested that a decrease in pain intensity of equal to or above 30% (pre-post treatment) is clinically meaningful.4,7 When changes in pain levels found in this study are

C EP

calculated as pre-post manipulation they correspond to a decrease of 37% in pain intensity. Thus, the changes in PC appear to lead to pain reductions which may be considered clinically meaningful.

4.3. Theoretical and clinical implications of the findings

The present study has several important theoretical and clinical implications. Theoretically, the finding that an increase in PC is associated with a heightened pain experience and a decrease in PC

C

is associated with a reduced pain experience in the same person provides strong evidence for a

A

direct association between the two phenomena. Also, the similar pattern of observations in both healthy volunteers and chronic pain patients further substantiates the findings and suggest similar effects and perhaps mechanisms in both intact and affected nociceptive systems. Furthermore, the prospective design enabled us to show that changes in PC may be prior or even causal to changes in pain ratings Based on previous findings, it is possible to speculate about how the level of PC may influence the experience of pain.6,11,16,19,20,25,33,34 PC is characterized by enhanced anticipation of pain, attention towards pain and negative emotional feelings in relation to pain.33 Brain imaging

21 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

studies have shown that PC is related to increased activity in the medial frontal cortex (anticipation), the dorsal anterior cingulate cortex, and the dorsolateral prefrontal cortex (attention to pain) as well as the rostral anterior cingulate cortex, the insula and the claustrum, closely related to the amygdala (emotional aspects of pain),11,25 processes which are known to facilitate nociceptive processing.19,20,34 Also, high levels of PC have been shown to be negatively related to activity in

D

prefrontal cortical regions involved in top-down modulation of pain such as the posterior cingulate and the parietal cortices.16,25 Thus, PC may influence the experience of pain both by facilitation of

TE

nociceptive processing and/or by inhibition of pain modulation, possibly primarily at the supra spinal level.32 Taken together, the findings of the present study seems to present the strongest – currently available – support for a direct relationship between PC and pain. However, the findings

C EP

do not preclude that in a different set-up manipulations of pain could also influence PC. In addition, as discussed in the method section, PC and pain are closely related in the PCS. Hence, one should be cautious when inferring causality.

Clinically, the findings are important because they show that not only healthy volunteers level of PC and pain but also chronic pain patients’ are influenced by the hypnotic suggestions and the magnitude of change appears to be within the range of clinical relevance (cf above). Thus, hypnotic

C

suggestions directly targeted at decreasing patients’ catastrophizing could be an effective and cost-

A

efficient intervention for chronic pain although long-term studies are needed to determine the actual clinical efficacy of this intervention. Furthermore, the finding that PC levels can be increased, apparently especially in chronic pain patients, points to the potential importance of carefully minimizing verbal and/or behavioral suggestions that may increase patients’ tendency to catastrophize about pain as this may in turn increase their pain levels.

4.4. Strengths and limitations of the study

22 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

Despite the strengths of the study outlined above, a potential limitation warrants consideration in the interpretation of the results. We tested the effect of the negative, neutral and positive hypnotic suggestions on only three of the thirteen items in the situation-specific PCS during the hypnotic trance so it can be speculated that the results may have been different if all thirteen items had been used. However, we were concerned that if participants had to rate all thirteen items during the

D

hypnotic trance it might have compromised the hypnotic trance and would have led to ethical and practical problems in maintaining a high level of experimental pain for healthy volunteers

TE

throughout the session (cf. method session). Still, we carefully selected the three items from the PCS that loaded most powerfully within each of the three subscales of the Danish version of the PCS.14 Also, we measured the dispositional PCS (all 13 items) pre and post hypnosis and found no

C EP

significant changes (see Table 4). These results suggest that the hypnotic suggestions specifically influenced the situation-specific PCS scores. Thus, we have no reason to believe that the results would have been different if all thirteen items had been used to assess the level of situation-specific PC during the hypnotic trance.

4.5. Conclusion

C

This is the first study to successfully manipulate pain catastrophizing in both positive and negative directions and observe subsequent substantial changes in pain intensity and unpleasantness in both

A

healthy volunteers and chronic pain patients. These results are of both theoretical and of clinical importance.

Acknowledgement We would like to thank statistician Bo Sommerlund and headache nurse Diana Schultz Knudsen. The study was supported by the MINDLab UNIK initiative at Aarhus University, which is funded by the Danish Ministry of Science, Technology and Innovation.

23 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

The authors report no conflict of interest. References

[1] Abrahamsen R, Dietz M, Lodahl S, Roepstorff A, Zachariae R, Østergaard L, Svensson P.Effect of hypnotic pain modulation on brain activity in patients with temporomandibular disorder

TE

D

pain. Pain 2010 Dec;151(3):825-33. doi: 10.1016/j.pain.2010.09.020. Epub 2010 Oct 8.

[2] Campbell C, McCauley L, Bounds S, Conn L, Simango M, Edwards R, Fontaine K. Changes in pain catastrophizing predict later changes in fibromyalgia clinical and experimental pain report:

C EP

Cross-lagged panel analyses of traditional and situational catastrophizing. J Pain 2012;13:3.

[3] Cohens J. Statistical power analyses for the behavioural sciences (second ed). 1988. Lawrance Erlbaum Associates.

[4] Dworkin RH, Turk DC, Wyrwich KW, Beaton D, Cleeland CS, Farrar JT, Haythornthwaite JA,

C

Jensen MP, Kerns RD, Ader DN. Interpreting the clinical importance of treatment outcomes in chronic pain clinical trials: IMMPACT recommendations. J Pain 2008;9:105-121.

A

[5] Edwards RR, Cahalan C, Mensing G, Smith M, Haythornthwaite JA. Pain, catastrophizing, and depression in the rheumatic diseases. Nat Rev Rheumatol 2011;7:216-224. [6] 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 2006;22:730-737. [7] Farrar JT, Young Jr JP, LaMoreaux L, Werth JL, Poole RM. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain 2001;94:149-158.

24 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

[8] Flor H, Turk DC. Chronic pain: An integrated biobehavioral approach. : IASP Press, 2011.

[9] Glisky ML, Kihlstrom JF. Hypnotizability and facets of openness. Int J Clin Exp Hypn 1993;41:112-123.

experience, and hypnotizability. J Pers Soc Psychol 1991;60:263.

D

[10] Glisky ML, Tataryn DJ, Tobias BA, Kihlstrom JF, McConkey KM. Absorption, openness to

[11] Gracely RH, Geisser ME, Giesecke T, Grant MAB, Petzke F, Williams DA, Clauw DJ. Pain

2004;127;4;853-843.

TE

catastrophizing and neural responses to pain among persons with fibromyalgia. Brain

C EP

[12] Grosen K, Vase L, Pilegaard HK, Pfeiffer-Jensen M, Drewes AM. Conditioned pain modulation and situational pain catastrophizing as preoperative predictors of pain following chest wall surgery: a prospective observational cohort study. PLoS One 2014;9:e90185.

[13] Katz J, Martin AL, Pagé MG, Calleri V. Alexithymia and fear of pain independently predict

C

heat pain intensity ratings among undergraduate university students. Pain Res Manag 2009;14:299.

A

[14] Kjøgx H, Zachariae R, Pfeiffer-Jensen M, Kasch H, Svensson P, Vase L, Jensen TS. Pain frequency moderates the relationship between pain catastrophizing and pain. Front Psychol 2014; [15] Kristiansen FL, Olesen AE, Brock C, Gazerani P, Petrini L, Mogil JS, Drewes AM. The Role

of Pain Catastrophizing in Experimental Pain Perception. Pain Practice 2014;14:E136-E145. [16] Lloyd D, Findlay G, Roberts N, Nurmikko T. Differences in low back pain behavior are reflected in the cerebral response to tactile stimulation of the lower back. Spine (Phila Pa 1976) 2008;33:1372-1377.

25 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

[17] Nijs J, Van de Putte K, Louckx F, Truijen S, De Meirleir K. Exercise performance and chronic pain in chronic fatigue syndrome: the role of pain catastrophizing. Pain Med 2007;9:1164-1172.

[18] Page RA, Handley GW, Carey JC. Can devices facilitate a hypnotic induction? Am J Clin

D

Hypn 2002;45:137-141.

TE

[19] Ploghaus A, Tracey I, Gati JS, Clare S, Menon RS, Matthews PM, Rawlins JN. Dissociating pain from its anticipation in the human brain. Science 1999;284:1979-1981.

[20] Porro CA, Baraldi P, Pagnoni G, Serafini M, Facchin P, Maieron M, Nichelli P. Does

C EP

anticipation of pain affect cortical nociceptive systems? J Neurosci 2002;22:3206-3214.

[21] Price DD, Bush FM, Long S, Harkins SW. A comparison of pain measurement characteristics of mechanical visual analogue and simple numerical rating scales. Pain 1994;56:217-226. [22] Rhudy JL, Martin SL, Terry EL, France CR, Bartley EJ, DelVentura JL, Kerr KL. Pain catastrophizing is related to temporal summation of pain but not temporal summation of the

A

C

nociceptive flexion reflex. Pain 2011; 152;794-801.

[23] Schiffman E, Ohrbach R, Truelove E, Look J, Anderson G, Goulet J, List T, Svensson P, Gonzalez Y, Lobbezoo F. Diagnostic criteria for temporomandibular disorders (DC/TMD) for clinical and research applications: recommendations of the International RDC/TMD Consortium Network and Orofacial Pain Special Interest Group. J Orofac Pain 2014; 28;1; 1-22. [24] Schmidt-Hansen PT, Svensson P, Jensen TS, Graven-Nielsen T, Bach FW. Patterns of experimentally induced pain in pericranial muscles. Cephalalgia 2006;26:568-577.

26 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

[25] Seminowicz DA, Davis KD. Cortical responses to pain in healthy individuals depends on pain catastrophizing. Pain 2006;120:297-306. [26] Severeijns R, van den Hout, Marcel A, Vlaeyen JW. The causal status of pain catastrophizing: an experimental test with healthy participants. Eur J Pain 2005;9:257-265. [27] Shor RE, Orne EC. Norms on the Harvard group scale of hypnotic susceptibility, form A.

D

Int J Clin Exp Hypn. 1963 Jan;11:39-47.

Psychol Assess 1995;7:524-532.

TE

[28] Sullivan MJL, Bishop SR, Pivik J. The pain catastrophizing scale: development and validation.

[29] Sullivan MJL, Thorn B, Haythornthwaite JA, Keefe F, Martin M, Bradley LA, Lefebvre JC.

C EP

Theoretical perspectives on the relation between catastrophizing and pain. Clin J Pain 2001;17:52.

[30] Svensson P, De Laat A, Graven-Nielsen T, Arendt-Nielsen L. Experimental jaw-muscle pain does not change heteronymous H-reflexes in the human temporalis muscle. Exp Brain Res 1998;121:311-318.

C

[31] Tellegen A. Brief manual for the differential personality questionnaire. Unpublished manuscript, University of Minnesota 1982.

A

[32] Terry EL, Kathryn A, Thompson, Rhudy JL. Experimental reduction of pain catastrophizing modulates pain report but not spinal nociception as verified by mediation analyses. Pain DOI: 10.1097/ j pain 0000000000000192 [33] Van Damme S, Crombez G, Bijttebier P, Goubert L, Van Houdenhove B. A confirmatory factor analysis of the Pain Catastrophizing Scale: invariant factor structure across clinical and nonclinical populations. Pain 2002;96:319-324.

27 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

[34] Vase L, Nikolajsen L, Christensen B, Egsgaard LL, Arendt-Nielsen L, Svensson P, Staehelin Jensen T. Cognitive-emotional sensitization contributes to wind-up-like pain in phantom limb pain patients. Pain 2011;152:157-162.

clinical descriptions and diagnostic guidelines. 1992.

D

[35] World Health Organization. The ICD-10 classification of mental and behavioural disorders:

TE

[36] Zachariae R, Sommerlund B, Molay F. Danish norms for the Harvard group scale of hypnotic susceptibility, form A. Int J Clin Exp Hypn 1996;44:140-152.

[37] Zachariae R, Jørgensen MM, Christensen S. Hypnotizability and absorption in a Danish

A

C

C EP

sample: testing the influence of context. Int J Clin Exp Hypn. 2000 Jul;48(3):306-14.

28 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

Figure legends Fig. 1: Study design for healthy volunteers and pain patients. HV- healthy volunteers Fig. 2: Change in the situation-specific PCS difference score (post-pre), in the Negative, Neutral and Positive Conditions for pain patients and healthy volunteers. In the Negative Condition, the situation-specific PCS difference score was positive, which indicates higher levels of pain

D

catastrophizing. In the Neutral and the Positive Condition, the situation-specific PCS difference

TE

score was negative, which indicates lower levels of pain catastrophizing

Fig. 3. Change in the NRS (intensity) difference score (post-pre), in the Negative, Neutral and Positive Conditions in pain patients and healthy volunteers. In the Negative Condition, the NRS difference score is positive, which indicates higher levels of pain. In the Positive and Neutral

C EP

Condition, the NRS difference score is negative, which indicates lower levels of pain

Fig. 4. Change in NRS (pain unpleasantness) difference score (post-pre), in the Negative, Neutral and Positive Conditions in pain patients and healthy volunteers. In the Negative Condition, the NRS difference score was positive, which indicates higher levels of pain. In the Positive and Neutral

A

C

Condition, the NRS difference score was Negative, which indicates lower levels of pain

29 Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

Table 1: Characteristics of participants Patients

Healthy volunteers

Mean/SD

Mean/SD

Males/Females

4/18

9/13

Age (years)

30/8.8

23.3/5.0

Dispositional Pain catastrophizing (PCS)

29.2/7.8

20.5/5.4

Pain Numeric Rating Scale (NRS)

5.7/2.6

0.1/0.5

Tellegen Absorption Scale (TAS)

18.00/6.29

20.00/4.24

A

C

C EP

TE

D

Characteristics of participants

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

Table 2: Regression analysis for situation-specific PCS difference score (post-pre) and NRS (pain intensity and unpleasantness) difference scores (post-pre) for patients and healthy volunteers Situation-specific pain catastrophizing in relation to pain intensity and pain unpleasantness.

Positive Condition

R2

p

Pain intensity

.327

.204

.045*

.430

.328

.005**

Pain unpleasantness

.377

.425

.002**

.617

.342

.004**

Pain intensity

.408

.304

.012*

.369

.251

.018*

Pain unpleasantness

.435

.302

.012*

.363

.174

.053

Pain intensity

.185

.036

.398

.699

.450

.001**

Pain unpleasantness

.256

.128

.777

.505

.000**

C EP

Neutral Condition

B

.112

B

R2

D

Negative Condition

Healthy volunteers

TE

Patients

p

*Correlation is significant at the 0.05 level

** Correlation is significant at the 0.01 level

A

C

B = unstandardized coefficients; R2 = R squared (coefficient of determination)

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

Table 3: Total increase and decrease for situation-specific PCS and NRS (intensity and unpleasantness) difference score (post-pre), in Negative and Positive Conditions compared to Neutral, in pain patients and healthy volunteers. An increase in score indicates higher levels of pain catastrophizing and pain, whereas a decrease in score indicates lower levels of pain catastrophizing and pain

Variable

PCS

Difference score in

Negative

vs. Pain intensity

Neutral Condition

Pain unpleasantness

Pain intensity

A

Positive vs. Neutral Condition

C

Difference scores (pre-post) in PCS

Pain unpleasantness

* Significant at the 0.05 level

Cohens d

C EP

M/SD

(pre-post)

Healthy

Difference

volunteers

between patients

TE D

Patients

and healthy volunteers

M/SD

Cohens d

p

3.45/2.48

1.41

1.86/1.67

1.11

.016*

1.28/2.11

0.62

1.36/1.76

0.77

.916

1.1/1.55

0.71

1.33/1.96

0.67

.690

-2.66/1.87

-1.44

-3.41/2.21

-1.54

.272

-1.62/1.93

-0.83

-2.27/1.88

-1.21

.260

-1.95/1.70

-1.15

-2.18/1.86

-1.17

.677

Cohens d is calculated using results from the Paired sample t-test

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

Table 4: Dispositional PCS measured before and after hypnotic suggestions Dispositional PCS measures: pre and post hypnotic suggestions

PCS pre hypnosis

Healthy volunteers

Mean/SD

Mean/SD

26.55/8.95

Condition

Positive

p = .215

p = .325

PCS post hypnosis

24.79/8.16

17.86/7.09

PCS pre hypnosis

26.25/7.38

17.55/5.65

Condition

p = .085 PCS post hypnosis

23.00/8.56

PCS pre hypnosis

25.64/8.55

Condition

p = .172

16.00/6.16

TE

Neutral

15.29/8.63

D

Negative

Patients

17.30/5.89

p = .137

PCS post hypnosis

15.05/5.55

A

C

C EP

* Significant at the 0.05 level

26.96/8.87

p = .119

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

D TE

A

C

C EP

Fig. 1

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

D TE C EP

A

C

Fig. 2

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

D TE

A

C

C EP

Fig. 3

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.

D TE

A

C

C EP

Fig. 4

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.