Escolar Documentos
Profissional Documentos
Cultura Documentos
www.elsevier.com/locate/pain
Clinical Psychology Unit, School of Psychology, University of Sydney, Australia Department of Physiotherapy, University of Sydney, Australia c Pain Management and Research Centre, Royal North Shore Hospital, Sydney, Australia
b
a r t i c l e
i n f o
a b s t r a c t
This study aimed to investigate the efcacy of an attention training technique (ATT) on pain ratings, threshold and tolerance during the cold pressor task. One hundred and three undergraduate students were randomly assigned to receive either threat-alleviating or threat-inducing information about the task. Participants were then re-randomized to receive either ATT or progressive muscle relaxation (PMR). Hence, the present study had a 2 (threat expectancy: high vs. low) 2 (training: ATT vs. PMR) design. Analyses conrmed that the threat manipulation was effective in increasing the harm associated with the task. ATT resulted in a relative reduction in hypervigilance to sensory pain words compared to PMR. ATT was also associated with a lower degree of focus on internal sensations, but not mindfulness or difculty disengaging from pain words. Results showed that, relative to relaxation training, those receiving ATT reported pain less quickly than those receiving relaxation, although there were no differences between the training groups for tolerance or pain ratings. These results show that ATT changes the cognitive processes of internal/external focus and hypervigilance towards sensory pain words, but not difculty disengaging or mindfulness. Although ATT changed threshold, the fact that neither pain ratings nor tolerance was affected suggests that a single, brief session of ATT may not be sufcient to affect broader change. Nonetheless, this study shows that ATT can change cognitive processes thought to be associated with heightened perception of pain and that this changes how quickly pain is registered and is therefore worthy of further investigation. Crown Copyright 2010 Published by Elsevier B.V. on behalf of International Association for the Study of Pain. All rights reserved.
Article history: Received 1 September 2009 Received in revised form 14 May 2010 Accepted 27 May 2010
1. Introduction Attention to painful stimuli has been highlighted as a determinant of response to pain [6]. If pain is interpreted as threatening, attention is directed to the pain rather than competing tasks. This prompts escape or avoidance behaviours to restore equilibrium [6]. There is considerable research that conrms that pain is associated with high attentional demands that are compounded when the threat value of pain is increased [25]. If attention to pain prompts avoidance, one might expect that interventions that reduce attention to pain would reduce pain and pain-related avoidance. Distraction, for example, has been used as a part of cognitivebehavioural therapy for patients with chronic pain [22]. However, research investigating distraction alone remains inconclusive [12]. It has been suggested that the efcacy of distraction may depend upon the threat value of pain [25]. Van Damme and colleagues manipulated threat and found
* Corresponding author. Address: School of Psychology A17, The University of Sydney, NSW 2006, Australia. Tel.: +61 2 9351 4558; fax: +61 2 9351 7328. E-mail addresses: louises@psych.usyd.edu.au (L. Sharpe), K.NicholsonPerry@ uws.edu.au (K. Nicholson Perry).
that distraction increased catastrophic cognitions and anxiety under conditions of high threat. They concluded that caution is required in using distraction when pain is highly threatening, as in clinical settings. Similar concerns have been discussed in the anxiety literature, with the result that distraction is now viewed as a form of safety behaviour [24]. As such, it is no longer seen as a useful adjunct to therapy with anxious patients. Instead other ways of changing attentional processes have been developed. Wells [28] has developed a method of attention training (ATT) that focuses on helping participants to effortfully re-direct their attention away from threatening internal stimuli, such as catastrophic cognitions thought to fuel anxiety. There is some preliminary evidence that this method is benecial in patients with panic disorders, social anxiety and hypochondriasis [23,29]. Wells argues that by shifting the focus of attention from internal threatening processes to external factors, anxiety is reduced [23,28,29]. Since pain, like anxiety, is a noxious and potentially threatening internal experience, the rationale for the efcacy of ATT can be expanded to pain. However, ATT has not been trialed in the pain literature. The aim of this study is to test the efcacy of ATT compared to a control intervention (i.e. relaxation training) using an acute
0304-3959/$36.00 Crown Copyright 2010 Published by Elsevier B.V. on behalf of International Association for the Study of Pain. All rights reserved. doi:10.1016/j.pain.2010.05.027
470
experimental pain task. We also aim to determine the likely mechanism of change. Participants were randomized to high or low threat conditions and re-randomized to receive ATT or relaxation before completing the cold pressor task. We hypothesized that the ATT group would have higher threshold and tolerance times and lower pain levels than the relaxation group. Measures were taken of possible cognitive processes that may be changed by ATT, including hypervigilance, difculty disengaging from pain words, mindfulness and internal/external focus to identify which cognitive processes were shifted as a result of ATT.
2.2.4. Toronto mindfulness scale [11] The Toronto mindfulness scale is a 13-item self-report measure with two sub-scales (curiosity and decentering) that assesses the construct of mindfulness (being aware of the present moment without judging or evaluating any thoughts that may arise) immediately after a meditation session. Internal consistency (a = 0.93, a = 0.91 for the curiosity and decentering sub-scales, respectively), criterion validity and discriminant validity are good [11]. For the present study, internal consistency was good for the total scale (a = 0.88) and the curiosity (a = 0.91) and decentering sub-scales (a = 0.72). 2.2.5. Self-attention manipulation check [30] The self-attention manipulation check is routinely employed following ATT and was used in the present study to assess participants self-focus after engaging in ATT and PMR [30]. It consists of a single item asking participants to rate the intensity of their self-focus on a 7-point numerical Likert scale. The Likert scale ranges from 3 (indicating they are entirely externally focused) and +3 (indicating they are entirely self-focused). 2.2.6. Pain measures Four pain measures were taken during the experiment. Threshold and tolerance times were recorded by the experimenter. Threshold refers to time (in seconds) at which the participants rst registered pain after placing their arm in the cold pressor. Tolerance refers to the time (in seconds) at which the participants withdrew their arm from the cold pressor. The experimenter also recorded participant pain ratings during the course of the task; that is, pain at threshold, pain 30 s after threshold and level of pain at tolerance. For these ratings, participants were asked to rate their pain on an 11-point numeric scale ranging from 0 (no pain) to 10 (extremely intense pain). 2.3. Materials 2.3.1. Dot-probe task [5,15] Attentional bias was assessed using a modied version of the dot-probe task, used extensively in previous pain research [5]. Subjects were presented with a xation point . that appeared in the centre of the computer screen for 500 ms. Following this, a pair of words was presented on screen for 500 ms, with one word presented above the xation point and one word presented below the xation point. The pair of words then disappeared from screen and one of these words was subsequently replaced by a probe (either the letter p or the letter q). Each participant was instructed to press either the letter p or the letter q on the keyboard, according to which had appeared on screen. A new pair of words was presented after a response was made or after 1500 ms had elapsed without a response. Participants completed ve practice trials with feedback from the experimenter, prior to completing the experimental trials to ensure that participants completely understood task requirements. Participants completed 200 experimental trials, which were presented in ve blocks of 40 trials, with a 1-min break between blocks. As such, the dot-probe task was approximately 12 min in duration. 2.3.2. Word stimuli The stimulus set for the dot-probe task was comprised of words from four pain-related threat categories (sensory, affective, disability and general) and neutral words. Each pain-related word pair contained a target word from one of the four pain categories and a matched neutral word, whilst word pairs in the neutral category contained two neutral words (i.e. neutralneutral trials). Within each of the word categories (sensory, affective, disability, general and neutral) there were 10 word pairs, matched in frequency and length. The pain-related word pairs were taken from previous
2. Method 2.1. Participants One hundred and four (38 males, 66 females) students from rst year psychology at the University of Sydney participated in the study for course credit. Exclusion criteria included a chronic pain condition or a condition that could be affected by induced pain or anxiety; any other current medical or psychological condition; recent use of analgesics; current signicant pain, rated as greater than 4/10 on a visual analogue scale; excessive caffeine or alcohol intake in the preceding 24 h (in accordance with NDARC and NHMRC guidelines, respectively); and inability to read or comprehend English. One participant was excluded from the study as they rated their current pain as 5/10 on a visual analogue scale and had consumed six standard units of alcohol in the preceding 24 h. No other participants were excluded. After excluding this participant, the nal sample contained 103 (38 males, 65 females) students, aged between 17 and 56 years (mean = 19.48, SD = 4.20). Participants were randomly assigned to one of the four experimental conditions (attention training/threat; attention training/no threat; progressive muscle relaxation/threat; progressive muscle relaxation/no threat) using the Research Randomizer program (www.randomizer.org). This study was approved by The University of Sydney Human Research Ethics Committee. 2.2. Measures 2.2.1. Fear of pain questionnaire (FPQ) III [19] The FPQ is a 30-item, self-report measure with three sub-scales (severe, minor and medical pain). It assesses the degree to which both pain and non-pain populations are fearful of pain. Internal consistency is excellent (a = 0.92) and testretest reliability is good (r = 0.74) [19]. In the present study, internal consistency was high for the total scale (a = 0.91) and for the severe (a = 0.85), minor (a = 0.83) and medical sub-scales (a = 0.88). 2.2.2. Depression, anxiety and stress scales (DASS) [14] The DASS is a 42-item self-report questionnaire with three subscales that assess depression, anxiety and stress. Internal consistency is high (a = 0.97) [4], convergent and discriminant validity are good, and psychometric properties are similar in clinical and non-clinical populations [3,4]. In the present study, internal consistency was high for the total scale (a = 0.95) and for the depression (a = 0.91), anxiety (a = 0.84) and stress (a = 0.90) sub-scales. [3,4]. 2.2.3. Threat manipulation check The manipulation check consisted of four questions that assessed the extent to which the cold pressor task information had induced threat or reassured the participant. The questions assessed worry, perceived harm, predicted pain level and coping ability associated with the task, with participants rating their answer on an 11-point numeric scale (010).
471
research [5] and the neutral word pairs were selected from previous research [8,9]. For each of the threat-related word pairs, four possible presentation types were used, with each of the word pairs presented once in each conguration (target-up/probe-up, targetup/probe-down, target-down/probe-up, and target-down/probedown). Thus, there were two main trial types: congruent trials (both the target word and probe appear in the same location) and incongruent trials (the target word and probe appear in different locations). Whilst there is no target in neutral word pairs, the same presentation congurations were still utilized, with one neutral word in each pair used in the same fashion as a target word. 2.3.3. Wells attention re-training task The attention training task (AT) was downloaded with permission from the Metacognitive Therapy Institute website (http:// mct-institute.com/, 20/05/2008) in an MP3 audio le format. Participants were asked to sit comfortably in their chair in front of the computer and read the following written introduction to the task: The attention training technique (ATT) is a technique that trains individuals to become more aware of their focus of attention and to help one gain control over their attention. ATT works by increasing your level of mental exibility by strengthening and restoring internal mental control mechanisms. It is normal that other inner thoughts and experiences will arise during ATT; these experiences should be treated as inner noises or passing events in your mind and body. Do not try to push them away or distract from them. Simply acknowledge their presence and then continue to focus your attention as instructed on the recording. Keep your eyes open for the duration of the recording. Do you have any questions before we begin? A written introduction to the task was utilized so that any questions could be answered by the experimenter before the task commenced. To avoid unnecessary repetition, the audio introduction was omitted. Throughout the ATT task, participants were directed through three attention tasks: selective attention, in which participants focused on one specic sound; switching attention, in which participants were required to focus on a specic sound and then another in rapid succession; and divided attention, in which participants were instructed to focus on as many different sounds as they could at once. In addition to being directed to focus their attention on sounds on the audio (e.g. bells, trafc, birds, and running water), participants were to focus their attention on sounds in the room that were behind them, to their left and to their right. Therefore, the volume of the audio was set at a moderate level so that participants could also attend to any sounds in the background. The humming of the cold pressor apparatus, a small clock radio playing classical music and the tapping of a pen by the experimenter served as background noises, and were at a volume comparable to the sounds on the audio. 2.3.4. Progressive muscle relaxation task A progressive muscle relaxation script (PMR) was developed by the researchers and recorded on a computer via a microphone attachment. Previous research has used PMR as a placebo-type intervention for comparison with other interventions (e.g. hypnosis) and has found no evidence that relaxation inuences threshold, tolerance or level of pain [7,20,21]. PMR was therefore thought to be an appropriate placebo-type control for therapeutic contact since it was unlikely to impact upon the attentional biases and may in fact represent a form of distraction or avoidance. The recording produced was a wave sound audio le format and was created using the program Gold Wave (http://www.goldwave. com/, 11/04/2008). Participants were asked to sit comfortably in their chair in front of the computer and read the following introduction to the task:
Progressive muscle relaxation (PMR) is a technique that trains individuals to successfully manage pain and stress, and in particular, to achieve a state of deep relaxation. PMR works on the principle that a muscle can be relaxed by rst tensing it for a few seconds and then releasing it. Tensing and relaxing various muscle groups throughout the body produces a state of relaxation. It is important that you maintain focus and listen carefully to the instructions on the recording in order to achieve a state of deep relaxation. You should be seated comfortably in your chair, with your feet at on the oor, legs uncrossed and your hands resting comfortably at your side or on your lap, with your eyes closed. Do you have any questions before we begin? After any questions were answered by the experimenter, the task began. During the task, participants were directed to tense various muscles throughout the body for 10 s each and then relax them, whilst focusing on their breathing. Participants were directed to tense their muscles in an ordered structure, beginning with the facial muscles and continuing slowly down the body, ending with the toes. 2.3.5. Cold pressor apparatus Two units manufactured by Thermoline Scientic Australia were used in the study to measure participant pain ratings and each unit consisted of three parts: a 20-L circulating water bath (model TLWB-30), an immersion cooler (TIC-400) and a heater circulator (TU3). Participants rst placed their right arm up to the elbow in the body temperature tank for 30 s, which was maintained at a temperature of 36 C (0.2 C) to ensure regulation of participants arm temperature. Participants then immediately placed their right arm up to the elbow in the cold tank which was maintained at a temperature of 5 C (0.2 C). Digital timers were set to turn the apparatus on and off at appropriate times, so that both water tanks were at the correct temperature for experimentation. This temperature was selected as it is within the optimum range to observe the Lewis effect, in which pain is experienced by the participant due to vasoconstriction and subsequent vasodilation of blood vessels [1,13]. 2.4. Procedure A diagrammatic representation of the procedure is provided in Fig. 1. Participants were brought into the laboratory for an individual testing session of approximately 1 h. All participants read an information statement that described the nature of the study and signed a consent form, with any queries being answered by the experimenter. After consent was obtained, participants completed a screening questionnaire to ensure that they fullled inclusion criteria, followed by the depression anxiety stress scales, the fear of pain questionnaire and the appropriate task information sheet, according to their experimental condition (threat/no threat). The information sheet described the task in one of the two ways, according to which grouping (threat or no threat) the participant had been randomized. For those in the threat group, the cold pressor task was referred to as a vasodilation task, and for those in the no threat group, the task was referred to as a cold pressor task. All information provided was accurate, however, participants in the threat group received information which emphasised the physical mechanisms of pain through the use of bio-medical terminology and an example of an extreme reaction to cold (frostbite) designed to induce fear. Participants in the no threat group received reassuring information in everyday terminology, which stressed the lack of danger associated with the task through the use of benign examples of exposure to the cold (searching for a cold drink in a bucket of ice). After reading the appropriate task information sheet, all participants completed a questionnaire which assessed the efcacy of the threat manipula-
472
Threat (n=51)
Non-threat (n=52)
Provided with study information Completed baseline questionnaires & manipulation check Dot-probe task I Explanation of task and practice trials Experimental trials
deal with the task. Participants then submerged their right arm up to the elbow in the body temperature tank for 30 s to control for baseline arm temperature. Participants then submerged the same arm in the cold pressor tank, withdrawing it at tolerance or at the maximum duration limit of 4 min. Pain threshold and tolerance times were recorded with a stopwatch, and participants verbal pain ratings were recorded manually by the experimenter at threshold, 30 s after threshold and again at tolerance. Pain ratings were recorded on an 11-point scale ranging from 0 (no pain) to 10 (extremely intense pain). After the participant removed their arm from the tank, they were reassured that there would be no long term damage from the task and that normal sensation would return to their arm shortly. All participants were then debriefed, verbally and in writing, and thanked for their participation. 2.5. Analyses
Dot-probe task II Experimental trials Cold pressor task Explanation of task and reminder to use strategy Arm submerged in body temperature tank for 30 s Arm submerged in cold water tank for maximum of 4 minutes (pain ratings taken at threshold, 30 s later and tolerance) Reassurance and debriefing
Fig. 1. Flow-chart of procedure for study. Key: ATT, attention training, PMR, progressive muscle relaxation.
tion. This manipulation was identical to threat manipulations reported in the literature [2,18]. Next participants were seated in front of the computer and given a verbal description of the word-based dot-probe task by the experimenter. The participant then read a written description of the task on screen, which was followed by several practice trials to ensure that the participants understood the task. The task was approximately 12 min in duration, and consisted of 200 trials, in which participants were presented with word pairs followed by a probe. Reaction times to the identication of the probe were measured to test for existing attentional pain biases. Participants then read a verbal description of either the Wells attention training task (ATT) or the progressive muscle relaxation task (PMR), before completing the task itself. It should be noted that the progressive muscle relaxation task served as the control, as it has previously been used as a placebo specically in comparison to ATT [17]. Both tasks were of equal duration, and lasted approximately 12 min each. Participants then completed two questionnaires, the Toronto mindfulness scale and self-awareness manipulation check, which assessed the effectiveness of the task, as well as participant task engagement. Following this, participants were again given the dot-probe task, in order to measure any change in attention towards pain as a result of the attention training task. Participants were then given a verbal introduction to the cold pressor task, in which the experimenter explained the procedure for the task, and how the threshold, tolerance and pain ratings would be taken. Any questions regarding the cold pressor task were answered by the experimenter. Participants were instructed to use the skill that they had been taught (i.e. ATT or PMR) to help
Based on the effect sizes achieved by Wells and colleagues in the anxiety literature [28,29], power analyses revealed a sample size of 26 participants in each group was necessary to achieve 80% power with a signicance level of 0.05. Initially, we performed a series of 2 (threat/no threat) 2 (training: ATT vs. relaxation) ANOVAs to determine whether there were any differences between the groups on any major variables. Any variables that differed signicantly between groups would be used as a covariate. We then performed two manipulation checks in order to ensure that each of the experimental manipulations had successfully manipulated the variables of interest. We performed a 2 (threat group) 2 (training group) on the manipulation check questions to ensure that the threat manipulation had increased the expectation of harm and a 2 (threat) 2 (training) ANOVA on Wells Likert item for use of ATT strategies. For cognitive processes, we assessed hypervigilance and difculty disengaging using the dot-probe paradigm. Individual reaction times for each participant on the dot-probe task were grouped according to word location (upper/lower) and dot probe location (upper/lower). In keeping with Keogh and colleagues [8,9], all response times less than 200 ms or greater than 1000 ms were removed as outliers. Inspection of subjects responses indicated that less than 5% of total responses were errors, and hence were not considered further in the analysis. In preparation for data analysis, averaged congruent and incongruent response latencies were calculated for each word pair using the following formulae [9]:
Averaged congruent response latencies tupu tlpl=2 Averaged incongruent response latencies tupl tlpu=2
In these formulae, t is the target, p is the probe, u is the upper location and l is the lower location. Thus, for example, tupl indicates that the target word appeared in the upper location of the screen and that the probe appeared in the lower location. Previous research has shown that the word groups of interest are sensory and affective pain words (e.g. [2,5]) and hence, we calculated the speed of responses to congruent and incongruent trials for these two word groups separately. We compared the averages for congruent and incongruent response latencies to the average for neutral-neutral trials [10], in a series of four mixed model ANOVAs. To determine hypervigilance, a 2 (word: experimental vs. neutral) 2 (time: pre- vs. post-training) 2 (threat: threat vs. no threat) 2 (training: ATT vs. PMR) ANOVA was conducted on responses to congruent trials for sensory and affective words. The same ANOVAs were also conducted based on response latencies to incongruent trials to assess difculty disengaging from the pain words. Measures of mindfulness and internal vs. external
473
focus were also assessed for differences between groups using a 2 (threat) 2(training) ANOVA, since these were administered only once following the training. The main analyses were 2 (threat group) 2 (training group) ANOVAs on the dependent variables derived from the cold pressor task. These variables were threshold and tolerance, plus pain at threshold and pain at tolerance. 3. Results 3.1. Participant characteristics A summary of participant characteristics is shown in Table 1. Chi-square tests were performed on dichotomous variables for the two independent variables: threat and training. These conrmed that there were no signicant differences between the groups for gender or handedness. A series of 2 (threat: threat vs. no threat) 2 (training: ATT vs. PMR) between group ANOVAs were conducted to test for baseline differences in normally distributed variables. No signicant between-group differences or interaction effects were found for age, fear of pain, current level of pain or any of the sub-scales or total score of the DASS. These results indicated that randomization had been successful. 3.2. Manipulation checks 3.2.1. (i) Effect of threat manipulation A 2 2 between-group ANOVA was conducted to verify that the cold pressor information elicited different responses from subjects allocated to the threat condition, compared to those in the non-threat group. Means are shown in Table 2. Results indicated that subjects in the threat group anticipated that the task could be signicantly more harmful (F(1,99) = 7.672, p = 0.007) compared to those in the non-threat group. Although there were no signicant differences between groups for how worried they were about the task, how much pain they thought it would cause or how likely they thought that they would be to cope with the task, there was a trend that the threat group expected to be less able to cope (F(1,99) = 3.355, p = 0.07). In addition, those in the ATT training group reported signicantly less worry than the PMR training group about the task (F(1,99) = 5.372, p = 0.023). The manipulation check questions were included to assess the efcacy of the threat manipulation and were administered prior to the intervention. Therefore, these differences reected an unintended group difference. As such, we controlled for level of worry in the nal analyses. There were no other signicant differences between training groups for any variables including dot-probe data.
Table 2 Mean participant ratings for manipulation check questions. Standard deviations in parentheses. Threat ATT Worry Pain Coping Harm* N
*
No threat PMR ATT (2.71) (2.84) (1.79) (2.16) 1.92 5.72 6.68 1.60 25 (2.04) (2.23) (1.35) (1.78) PMR 3.08 5.65 5.96 1.12 26 (2.12) (2.33) (1.99) (1.42)
Indicates a signicant difference between the two threat groups (p < 0.05).
3.2.2. (ii) Effect of training A 2 (training: ATT/PMR) 2 (threat: threat/no threat) ANOVA was conducted to examine the effect of training on measures of attentional focus. This demonstrated a signicant main effect of training on direction of focus (internal vs. external), with those receiving the ATT training reporting more neutral focus compared to more internal focus in the PMR group (F(1,99) = 7.081, p = 0.009). There were no differences between the training groups on the measure of mindfulness, and no differences on any measure between the threat groups. 3.2.3. (iii) Cognitive processes Two sets of mixed model ANOVAs were conducted to test for the effects of training on hypervigilance and disengagement. Two 2 (training: ATT/PMR) 2 (word: congruent experimental/neutral) 2 (time: pre/post-training) mixed model ANOVAs were conducted, using sensory and affective pain words to test for effects of hypervigilance. This indicated that there was an interaction effect between training and time for hypervigilance to sensory words (F(1,96) = 4.701, p = 0.033) but not to affective ones (F(1,96) = 2.876, p = 0.093). This interaction effect indicates that the ATT group changed in such a way that they became less hypervigilant towards sensory pain words, while, at the same time, the PMR group changed in the opposite direction (see Table 3 and Fig. 2). Two 2 (training: ATT/PMR) 2 (word: incongruent experimental/neutral) 2 (time: pre/post-training) mixed model ANOVAs were conducted, using sensory and affective pain words to test for effects of disengagement. There were no signicant effects detected for disengagement. 3.3. Main outcome analyses A series of 2 (training: ATT/PMR) 2 (threat: no threat/threat) ANCOVAs were conducted (controlling for harm associated with the task) for threshold, tolerance and pain levels at 30 s and at tol-
Table 1 Mean baseline participant characteristics by threat (threat/no threat) and training (ATT/PMR) groups. Standard deviations in parentheses. Variables Threat ATT Gender Age FPQ-III DASS_D DASS_A DASS_S DASS Tot N 16 females 18.69 (1.19) 82.42 (17.91) 3.12 (3.08) 3.65 (3.32) 7.50 (3.89) 14.27 (7.91) 26 PMR 15 females 19.04 (1.59) 84.38 (17.22) 5.31 (7.13) 5.12 (5.28) 8.73 (6.92) 19.15 (16.85) 26 No threat ATT 15 females 21.20 (7.84) 79.76 (13.20) 4.40 (6.40) 4.24 (6.04) 9.20 (8.09) 17.84 (19.32) 25 PMR 19 females 19.04 (2.25) 81.15 (17.73) 4.00 (3.83) 3.81 (3.80) 9.00 (6.49) 16.81 (12.31) 26
Table 3 Mean reaction times (ms) for congruent and incongruent trials for the ATT and PMR groups pre- and post-training for sensory pain words. Standard deviations in parentheses. ATT Congruent Pre-treatment Post-treatment 480.45 (68.58) 425.91 (46.33) Pre 5.06 52 Incongruent 473.58 (66.73) 428.96 (49.37) Post 5.44 52 PMR Congruent 497.37 (88.97) 448.24 (81.85) Pre 5.62 52 Incongruent 506.95 (90.94) 444.14 (82.07) Post 1.37 52
Congruence index* N
FPQ-III, fear of pain questionnaire-III; DASS, depression anxiety stress scale; D, depression; A, anxiety; S, stress; Tot, total; N, number of participants in each group. There were no signicant differences between the groups on any variable.
ATT, attention training technique; PMR, progressive muscular relaxation. * The congruence index is calculated by subtracting the averaged response latencies of congruent trials from the average response latency of neutral/neutral trials.
474
ATT PMR
Fig. 2. Training (ATT vs. PMR) by Time (Pre vs. Post) interaction for sensory words.
Table 4 Mean participant scores for threshold, tolerance, threshold pain at 30 s and pain at tolerance. Standard deviations in parentheses. Threat ATT Threshold 10.77 (8.30) (seconds) Pain at 30 s 4.88 (2.07) Tolerance (seconds) 66.92 (56.20) Pain at tolerance 8.23 (2.03) N 26 PMR 7.69 (7.57) 5.35 (1.72) 54.27 (50.01) 8.54 (1.48) 26 No threat ATT 12.12 (13.05) 4.64 (1.66) 84.36 (80.02) 8.24 (1.36) 25 PMR 8.00 (4.95) 4.46 (1.77) 107.50 (99.39) 7.69 (2.02) 26
erance. For threshold, there was a main effect for training (F(1,99) = 4.070, p = 0.046), but not for threat (F(1,99) = 0.232, p = 0.631). There was no interaction effect for threshold (F(1,99) = 0.090, p = 0.765). Examination of the means presented in Table 4 indicates that those who received PMR training registered pain more quickly than those in the ATT training condition. In contrast, for tolerance, there was a main effect for threat (F(1,99) = 5.172, p = 0.025) but not training (F(1,99) = 0.480, p = 0.490), and there were no interaction effects (F(2,99) = 1.641, p = 0.203). Table 4 indicates that those who were in the threat condition withdrew their arm more quickly than those in the reassuring information condition. There were no signicant differences in pain ratings at 30 s or tolerance between those in different threat groups or training groups, and neither were there any interaction effects. 4. Discussion The present study aimed to compare the efcacy of ATT with relaxation on threshold, tolerance and pain ratings during a cold pressor task under high and low threat expectancy conditions. We hypothesized that ATT would result in increased tolerance and threshold and lower pain ratings. These hypotheses were supported only for threshold, with participants in the ATT condition
reporting pain later during the cold pressor task than those receiving relaxation. ATT aims to help patients shift attention from threatening internal sensations to external stimuli, hence we measured key cognitive processes likely to be effected by ATT, including external/internal focus, hypervigilance, difculty disengaging and general mindfulness. Relative to the PMR group, ATT resulted in signicant changes over time in hypervigilance to sensory, but not affective pain words, and also led participants to be more externally focused in their attention, as the intervention aims to do. Before considering these results in detail, there were a number of limitations that warrant discussion. Firstly, participants in this study were undergraduate students who took part in the cold pressor task. Hence, the degree to which these results are generalizable to those with acute or chronic pain injuries is speculative. Clearly, the potential importance of ATT relates to its ability to inuence pain in real-life settings and clearly making inferences about such effects would be premature. Secondly, while ATT increased threshold this does not necessarily indicate that participants experienced the pain earlier. Earlier reports of pain could be due to response bias. It is also possible that the inclusion of two experimental manipulations (i.e. threat and intervention) complicated the design and weakened the effect of either one or both interventions. Finally, the use of relaxation as a control condition was included because we felt that it was important to include a placebo-type condition, which seemed plausible to participants and has been used in previous research using ATT [17]. The fact that relaxation also includes a component of distraction is potentially a strength of this study and provides a potentially rigorous test of the efcacy of ATT by controlling for time, attention and the use of distracting strategies. However, it remains possible that the primary difference between these interventions is that relaxation causes individuals to focus internally, whereas ATT causes an external focus, the latter of which is more effective on threshold. The fact that we did not use a no intervention group means that we cannot be clear whether ATT increased threshold or relaxation reduced threshold. Even if one assumes that ATT had a benecial effect on the initial reporting of pain, it remains disappointing that other indices of outcome (namely tolerance and pain ratings) failed to differentiate between the training groups. The failure to nd an effect of other attention-based interventions on pain ratings is not an uncommon nding and may have been inuenced by the fact that these were only assessed on two occasions. Boston & Sharpe [2] trained participants to either attend to sensory or affective pain stimuli. In this study, the strategys effectiveness was dependent on the threat condition, with sensory training being relatively helpful in low threat but detrimental in high threat conditions on the outcomes of threshold and tolerance, but not on pain ratings. Van Damme et al. [25]s study was not designed to measure tolerance time, but they found that fewer participants withdrew their hand from the cold pressor task in the distraction condition (an indication of avoidance), despite no differences in pain ratings during the task. Interestingly, in their study, pain ratings after the task were affected by distraction. Unfortunately, we did not measure pain after the procedure and so whether ATT had similar effects is unknown. While Van Damme et al. [25] have suggested that distracting interventions may exhibit their effect on behavioural rather than pain outcomes, it is notable that we only found an effect of threat and not ATT on tolerance. McGowan et al. [18] who examined cognitive bias modication also failed to nd effects on tolerance, despite nding an effect of that intervention on threshold and pain at threshold. They suggested that it may be that there is an effect of the re-training, however, that this effect is short-lived affecting only proximal outcomes. That is one potential explanation also for our ndings. It may be that ATT effects behavioural responses
475
and not the experience of pain, however, because the intervention was designed to be extremely brief (12 min) that the effect is short-lived. It is also possible that the specic effects of the intervention (i.e. on threshold and not other outcomes) are related to the specic nature of the cognitive effect of ATT. In this study, we measured a number of potential mechanisms of change. The results clearly showed that ATT did affect some cognitive processes and not others. Consistent with its primary aim [28,29], ATT changed the internal/external focus of attention such that participants who completed ATT were more externally focused. Similarly, there was a change in hypervigilance to sensory pain words suggesting that after training, participants who received ATT became less vigilant to pain words compared to their baseline levels and to neutral words. These effects make sense, since ATT specically aims to reduce focus on internal processes and sensory pain words are the lexical representations of the sensations that participants might expect to experience during the cold pressor task. There was no evidence, from the dot-probe that ATT changed difculty disengaging from pain-related stimuli or changed hypervigilance to affective pain words. Nor was there evidence that the intervention changed general mindfulness, such as curiosity or ability to decenter. The failure to change these variables may have reduced the efcacy of the intervention [16]. Hypervigilance to pain describes the degree to which individuals scan their internal environment and notice sensations of pain. Therefore it makes sense that, if participants are trained to attend less to internal sensations, hypervigilance, rather than other cognitive processes will be affected. Similarly, threshold is to some degree the behavioural operationalisation of hypervigilance. That is, how long does it take participants to notice and report pain. It is intuitively sensible therefore that threshold may be the outcome variable most affected by changes to hypervigilance and an internal/external attention focus, as was found. Van Damme and colleagues [26,27] in two experiments found that initial shifting of attention to pain (hypervigilance as measured in this study) was not responsible for the attention to pain often reported in the pain literature. Rather, they found that difculty disengaging from painrelated stimuli was what characterised attentional biases to pain. Therefore, although ATT may produce shifts in initial orienting towards pain words and reduce internal focus, which may in turn result in lengthened threshold, it may be its failure to address cognitive components of difculty disengaging which limits its efcacy. Future research needs to explore other interventions that target attentional processes in pain and ensure that they measure different aspects of the attentional processes to help isolate those processes that are most strongly associated with pain responses. For example, mindfulness is an intervention which has gained some popularity as a treatment of pain in recent years [16]. Mindfulness aims not to change the focus of attention, but rather the meaning of the stimuli. That is, people are encouraged to experience the sensation objectively without positive or negative interpretation. By doing so, negative sensations are seen to lose their valence and hence be less salient and distressing. Future research could replicate the present design using a mindfulness-based attentional training. In summary, this study brings together a number of important ndings in the area of pain and attention. First, we were able to show that ATT can produce changes in the attentional processes of internal/external focus and hypervigilance to sensory pain words. This is rst demonstration of the effect of ATT on an experimental paradigm, such as the dot-probe. Further, there was an effect of ATT compared to relaxation on threshold that suggests that those receiving ATT report pain later in the task. This demonstrates clearly that manipulating the focus of attention can directly impact on individuals response to a painful task in ways predicted by the
pain and attention model [6]. While these results are encouraging, the main effects of training were observed only on one of the four indices of outcome, hence the limited nature of gains questions the clinical applicability of ATT to pain at this point in time. Future research should work towards developing interventions that could be trialled in real-life settings, such as with acute pain patients and in settings where individuals experience acute pain to determine whether these results can generalize to real-life applications. Additionally, it would be worthwhile to determine whether adding attention-based strategies to already effective treatments for chronic pain problems (e.g. pain management) facilitated the efcacy of those treatments. It is essential that future research clearly measures the cognitive processes that are affected by interventions based on manipulating attention to help better understand the role of attention in pain response. Conicts of interest The authors have no conicts of interest. Acknowledgements This study was funded by the Australian Research Council in a Discovery Project (DP0773250) to L.S., M.K.N. and K.R. L.S. is supported by an NHMRC Senior Research Fellowship. References
[1] Ahles TA, Blanchard EB, Leventhal H. Cognitive control of pain: attention to the sensory aspects of the cold pressor stimulus. Cognitive Ther Res 1983;7:15978. [2] Boston A, Sharpe L. The role of threat-expectancy in acute pain: effects on attentional bias, coping strategy effectiveness and response to pain. Pain 2005;119:16875. [3] Brown TA, Chorpita BF, Korotitsch W, Barlow DH. Psychometric properties of the depression anxiety stress scales (DASS) in clinical samples. Behav Res Ther 1997;35:7989. [4] Crawford JR, Henry JD. The depression anxiety stress scales (DASS): normative data and latent structure in a large non-clinical sample. Br J Clin Psychol 2003;42:11131. [5] Dehghani M, Sharpe L, Nicholas MK. Selective attention to pain-related information in chronic musculoskeletal pain patients. Pain 2003;105:3746. [6] Eccleston C, Crombez G. Pain demands attention: a cognitive-affective model of the interruptive function of pain. Psychol Bull 1999;125:35666. [7] Houle M, McGrath PA, Moran G, Garrett OJ. The efcacy of hypnosis- and relaxation-induced analgesia on two dimensions of pain for cold pressor and electrical tooth pulp stimulation. Pain 1988;33:24151. [8] Keogh E, Ellery D, Hunt C, Hannent I. Selective attentional bias for pain-related stimuli amongst pain fearful individuals. Pain 2001;91:91100. [9] Keogh E, Dillon C, Georgiou G, Hunt C. Selective attentional biases for physical threat in physical anxiety sensitivity. J Anxiety Disord 2001;15:299315. [10] Koster EHW, Crombez G, Verscuere B, De Houwer J. Selective attention to threat in the dot probe paradigm: differentiating vigilance and difculty to disengage. Behav Res Ther 2004;42:118392. [11] Lau MA, Bishop SR, Segal ZV, Buis T, Anderson ND, Carlson L, Shapiro S, Carmody J, Abbey S, Devins G. The Toronto mindfulness scale: development and validation. J Clin Psychol 2006;12:144567. [12] Leventhal H. I know distraction works even though it doesnt. Health Psychol 1992;11:2089. [13] Lovallo W. The cold pressor test and autonomic function: a review and integration. Psychophysiology 1975;12:26882. [14] Lovibond SH, Lovibond PF. Manual for the depression anxiety stress scales. Sydney, NSW: Psychology Foundation; 1995. [15] MacLeod C, Mathews A, Tata P. Attentional bias in emotional disorders. J Abnorm Psychol 1986;95:1520. [16] McCracken LM, Gauntlett-Gilbert J, Vowles KE. The role of mindfulness in a contextual cognitive-behavioral analysis of chronic pain-related suffering and disability. Pain 2007;7:639. [17] McEvoy PM, Perini SJ. Cognitive behavioral group therapy for social phobia with or without attention training: a controlled trial. J Anxiety Disord 2009;23:51928. [18] McGowan N, Sharpe L, Refshauge K, Nicholas M. The effect of attentional retraining and threat expectancy in response to acute pain. Pain 2009;142:427. [19] McNeil DW, Rainwater AJ. Development of the fear of pain questionnaire III. J Behav Med 1998;21:389410. [20] Miller MF, Barabasz AF, Barabasz M. Effects of active alert and relaxation hypnotic inductions on cold pressor pain. J Abnorm Psychol 1991;100:2236.
476
L. Sharpe et al. / PAIN 150 (2010) 469476 [26] Van Damme S, Crombez G, Eccleston C. Retarded disengagement from pain cues: the effects of pain catastrophizing and pain expectancy. Pain 2002;100:1118. [27] Van Damme S, Crombez G, Eccleston C. The anticipation of pain modulates spatial attention: evidence for pain-specicity in high-pain catastrophizers. Pain 2004;111:3929. [28] Wells A. Panic disorder in association with relaxation induced anxiety: An attentional training approach to treatment. Behav Ther 1990;21:27380. [29] Wells A, White J, Carter K. Attention training: effects on anxiety and beliefs in panic and social phobia. Clin Psychol Psychother 1997;4:22632. [30] Wells A. Emotional disorders and metacognition. Chichester, UK: Wiley; 2000.
[21] Mills WW, Farrow JT. The transcendental meditation technique and acute experimental pain. Psychosom Med 1981;43:15764. [22] Morley S, Eccleston C, Williams A. Systematic review and meta-analysis of randomized controlled trials of cognitive behaviour therapy and behaviour therapy for chronic pain in adults, excluding headache. Pain 1999;80:113. [23] Papageorgiou C, Wells A. Effects of attention training on hypochondriasis: a brief case series. Psychol Med 1998;28:193200. [24] Salkovskis PM, Clark DM, Hackmann A, Wells A, Gelder MG. An experimental investigation of the role of safety-seeking behaviours in the maintenance of panic disorder with agoraphobia. Behav Res Ther 1999;37:55974. [25] Van Damme S, Crombez G, Van Nieuwenborgh-De Wever K, Goubert L. Is distraction less effective when pain is threatening? An experimental investigation with the cold pressor task. Eur J Pain 2008;12:607.