Increased bias to report heat or pain following emotional priming of pain-related fear by Stuart Derbyshire | Papers by Stuart

Pain 137 (2008) 60–65 www.elsevier.com/locate/pain Increased bias to report heat or pain following emotional priming of pain-related fear S.S. Kirwilliam, S.W.G. Derbyshire * University of Birmingham, School of Psychology, Edgbaston B15 2TT, UK Received 10 April 2007; received in revised form 6 August 2007; accepted 10 August 2007 Abstract Emotional and attentional factors have been identified to play a significant role in modulating pain perception with negative emotions increasing pain sensitivity. Recent studies suggest that fearful images may activate the attentional components of fear driven behaviours and facilitate an attentional bias or sensitivity toward noxious stimuli. The current investigation examines whether priming of pain-related fear will affect performance by increasing sensitivity to punctuate heat stimuli. A modified version of the visual dot probe task was employed to provide priming of pain-related fear and a heat detection task was used to measure the effects of priming on sensitivity. The results indicated a significant facilitation of heat and pain perception at varying temperatures following emotional priming. In particular, there was an increase in the bias toward reporting a heat stimulus following emotional priming. The findings emphasise the efficacy of the visual dot probe task as a method of priming and provide a possible method for probing hypervigilance in chronic pain patients. Ó 2007 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. Keywords: Nociception; Somatic; Attention; Catastrophising; Hypervigilance; Chronic pain 1. Introduction The cerebral activity measured in many pain investigations is readily attributed to neurophysiologic sensory input. The perception of pain, however, is a subjective experience that can be influenced by affect and previous memories (Peyron et al., 1999). Current research has subsequently focused on the role of cognitive and emotional processes as modulators of pain response (e.g. Fulbright et al., 2001; Kenntner-Mabiala and Pauli, 2005; Wicch et al., 2005). The role of attention in modulating the perception of pain is well established (Eccleston and Crombez, 1999). Many studies have reported subjective ratings of pain to reduce during performance of an attenCorresponding author. Tel.: +44 121 414 4659; fax: +44 121 414 4897. E-mail address: s.w.derbyshire@bham.ac.uk (S.W.G. Derbyshire). * tion-demanding task (e.g. Good et al., 1999). Conversely, intense sensations of pain are reported when attention is focused toward a noxious stimulus (Miron et al., 1989). If selective attention is actively diverted elsewhere the level of attentional processing associated with pain appears to reduce. In contrast, the experience of pain can be intensified by increasing the degree of attentional processing focused on painful stimulation. It is also well documented that emotion can modulate the experience of pain (Villemure and Bushnell, 2002). For example, Wied and Verbaten (2001) exposed subjects to pleasant, unpleasant or neutral pictures, while administering the cold-pressor test. The results indicated tolerance to pain as a function of picture pleasantness, emphasising the beneficial and detrimental affect of positive and negative emotions, respectively. It has been argued that emotional processes modulate the perception of pain via priming. The motiva- 0304-3959/$34.00 Ó 2007 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2007.08.012 S.S. Kirwilliam, S.W.G. Derbyshire / Pain 137 (2008) 60–65 61 tion priming hypothesis (Lang, 1995) views emotions to be governed by two opponent behavioural systems, the appetitive (approach) and aversive (avoidance) systems. Furthermore, the motivational priming hypothesis also predicts that negative emotions increase sensitivity toward pain via activation of the aversive system. This prediction has received empirical support. In a study investigating the role of pain-related negative emotions, as opposed to negative emotion per se, Janssen et al. (1998) primed subjects with either painrelevant or pain-irrelevant anxiety. The results indicated a significantly greater level of sensitivity toward painful stimulation following priming of pain-relevant anxiety. Priming with fearful images has also been demonstrated to reduce cold-pressor pain tolerance (Meagher et al., 2001). Unlike previous studies, the present investigation will aim to achieve emotional priming of pain-related fear using a modified version of the visual dot probe task. This task is commonly employed to measure biases in attention in clinical disorders, such as anxiety (Bradley et al., 1999), and there is evidence to suggest that the visual dot probe task provokes activation of selective attention (Larsson et al., 2006). Imaging studies have demonstrated that selective attention excites cortical structures associated with the pain matrix (Bantick et al., 2002). Thus, we expect that the visual dot probe in conjunction with pain-relevant images will be an effective method of achieving emotional priming. The present study examines the capacity of primed pain-related fear as a modulating factor of subjective sensitivity to a heat detection task. In addition, the study will assess whether priming of pain-related fear affects performance by increasing the sensitivity and/ or bias toward heat and pain detection. 2. Methods 2.1. Subjects Fifty healthy (19 males, 31 females) right-handed subjects from the University of Birmingham were recruited via e-mail and poster advertisement and provided written informed consent. The study was approved by the Local Ethics Committee. The sample was divided into four groups and two experiments. Experiment 1: Each group consisted of 10 students who were assigned randomly to either the emotionally primed or control condition. Primed and control subjects had a mean age of 21.1 (SD = 1.7) and 21.3 (1.5) years, respectively (difference not significant). Experiment 2: Each group consisted of 15 students who were also assigned randomly to either the emotionally primed or control condition. Primed and control subjects had a mean age of 21.9 (1.3) and 22.4 (1.6) years, respectively (difference not significant). None of the subjects in either experiment had been administered any analgesics (e.g. paracetamol) prior to testing, had previously been diagnosed with any pain or affect disorder, or suffered from visual impairments that may potentially impede performance. 2.2. Apparatus A modified version of the visual dot probe task was programmed using SuperLab Pro. version 2.0 (CedrusÒ). The program was run on a Windows XP driven laptop computer and visual stimuli were presented on a 14-in. VGA monitor. Phasic heat stimuli were delivered via a circular thermode with a 27 mm surface diameter using the CHEPS NeuroSensory System (Medoc Advanced Medical systems; Ramat Yishia, Israel). This device allowed delivery of focal and extremely rapid pulses of heat. 2.3. Design A one-way between-subjects design was used to compare group heat detection performance and mask the intention of the study. The independent variable of image content consisted of two levels, pain-related fear and neutral. The dependent variable of sensitivity to heat detection was measured by performance on a heat detection task. Target stimuli were present in only half of the trials of the detection task. The design enabled the comparison of responses when subjects, (i) perceived a stimulus as heat (heat detection hit), (ii) perceived a stimulus as painful (pain detection hit) or (iii) did not perceive a stimulus (detection miss) during trials with and without heat stimulation. In addition, our design provided the necessary responses for signal detection theory (SDT) calculations. Such measurements assess whether priming of pain-related fear affects performance by increasing the sensitivity and/or bias toward heat and pain detection (Stanislav and Todorov, 1999). Experiment 1: Heat stimuli of two baseline temperatures (34 and 42 °C) were delivered to the volar surface of either forearm (left and right). The baseline temperature and lateral forearm of initial testing were counterbalanced among subjects to control for pain adaptation. Phasic heat was created by adding heat pulses that increased to 0.5, 1.0 and 1.5 °C above baseline. Each target temperature was administered six times (2 · 10 ms, 2 · 50 ms and 2 · 100 ms). Varying stimulus durations were applied in order to control for pain adaptation at different temperatures (Koyama et al., 2004). The task also involved an equal number of trials that did not deliver a stimulus. In total, each baseline temperature condition consisted of 36 trials, which were randomised to control for practice effects and counterbalanced across subjects to control for order effects. Table 1 illustrates six possible outcomes of this design used to measure sensitivity to the detection of heat. Experiment 2: Heat stimuli from a single baseline (37 °C) were delivered to the volar surface of the right forearm. Phasic heat was created by adding heat pulses that increased by 1 °C above baseline and lasted for 100 ms. To control for pain adaptation, the probe was rotated around three regions of the forearm after every pulse or no pulse trial. Sixty pulses were delivered to each subject in five blocks containing 12 62 S.S. Kirwilliam, S.W.G. Derbyshire / Pain 137 (2008) 60–65 Table 1 Possible outcome of the heat detection task Response Stimulus Pulse Pain Heat No stimulus Pain detection hit Heat detection hit Detection miss No pulse False pain hit False heat hit Correct rejection pulses with rests in between blocks. The task also involved an equal number of trials that did not deliver a stimulus to provide 120 detection trials with the possible outcomes as shown in Table 1. 2.4. Procedure Subjects completed the visual probe and heat detection tasks in a dimly lit quiet room, in order to reduce distractions and facilitate the performance of selective attention. 2.4.1. Visual probe task The visual probe task was employed to achieve emotional priming effects by accessing selective attention. Experimental pictures contained fearful images related to human pain (for example, a female face with bruising around the eye, a male clutching his abdomen). Control pictures excluded fearful images and the suggestion of pain. Subjects were then seated 60 cm in front of the computer screen that displayed a description and the instructions for the visual probe task. Each trial began with a central fixation point (+) presented for 1000 ms. Following offset of the fixation point a pair of pictures was presented side-by-side for 500 ms. Immediately after the pictures disappeared a dot probe (•) replaced the position of one of the preceding pictures. The probe remained on the screen until the subject indicated detection of the probe via a response key. The inter-trial interval lasted for 750 ms after which the next trial automatically initiated. Response times were not calculated as the visual probe task was only employed as a method of emotional priming via visual selective attention. The visual probe task began with practice trials, containing neutral images, followed by either control or experimental trials, which consisted of neutral images or fearful images related to pain, respectively. Visual stimuli were presented in a randomised order according to SuperLab Pro. Control trials contained two neutral images and experimental trials contained one pain-related fear and one neutral image. Practice trials incorporated 10 pairs of pictures presented twice while neutral and pain-related fear priming conditions consisted of 10 pairs of pictures presented four times. This design ensured an equal number of trials presenting each picture of every pair on the left and right side of the screen. This was intended to control for structural hemispheric differences in accessing selective attention. 2.4.2. Heat detection task Experiment 1: The CHEPS thermode was strapped to the right or left volar forearm and a series of heat pulses were delivered every 10 seconds. The subjects were cued for a subjective response every 5 seconds. The subjects were not aware of the delivery pattern but were informed that stimuli would sometimes be delivered and sometimes not. If the subject did not perceive a stimulus they were asked to respond by stating ‘no stimulus’. In the event that a stimulus was perceived the subject was asked to specify if the stimulus was ‘hot’ or ‘painfully hot’. These responses were recorded every 5 seconds until the end of the detection task. At the end of testing at one baseline temperature condition the thermode was removed and attached to the opposite forearm. The detection task was then repeated as before but using the other baseline temperature. Experiment 2: The CHEPS thermode was held against the right volar forearm by an experimenter. As for Experiment 1, a series of heat pulses were delivered every 10 seconds and subjects asked to report their experience every 5 seconds. The thermode was moved after each reported experience and rotated around three spots on the volar forearm. After 24 detection trials (involving 12 actual heat pulses) the probe was removed and the subject rested for 2.5 min before starting again. This was continued for 600 detection trials separated into five data collection blocks. 2.5. Data analysis Nonparametric v2 tests were used to examine the data recorded as shown in Table 1 in the primed and neutral groups. Signal detection theory (SDT) (Stanislav and Todorov, 1999) was used to calculate the ‘hit rate’, ‘false alarm rate’, sensitivity (d 0 ) and response bias (c) and to provide parametric data for analysis. Following the procedures described by Stanislav and Todorov (1999), we calculated the hit rate by dividing the number of hits by the total number of signal trials. Thus, in our experiment hit rate was calculated by dividing the number of pain and heat responses (when a signal was present) by the hits plus misses (i.e., the total number of signals delivered regardless of detection). Similarly, we calculated the false alarm rate by dividing the number of false alarms by the total number of noise trials. Thus, in our experiment false alarm rate was calculated by dividing the number of false alarms (i.e., pain or heat responses when no signal was present) by the false alarms plus correct rejections (i.e., the total number of noise trials regardless of detection). d 0 was calculated by subtracting the z score corresponding to the false alarm rate from the z score corresponding to the hit rate. c was calculated by averaging the z score that corresponds to the hit rate and the z score that corresponds to the false alarm rate and multiplying the result by negative one. These calculations were carried out in Excel using standard formulae described elsewhere (Stanislav and Todorov, 1999). These SDT calculations were entered into a series of repeated measures (Experiment 1) or one-way (Experiment 2) ANOVAs for formal comparison of the primed and neutral groups. 3. Results 3.1. Experiment 1 Table 2 summarises the subject responses to the heat pulses generated from a 34 °C baseline and Table 3 summarises the responses from a 42 °C baseline. A series of S.S. Kirwilliam, S.W.G. Derbyshire / Pain 137 (2008) 60–65 Table 2 Total of occurrences for each possible outcome of the heat detection task during 34 °C baseline temperature trials Outcome Group Heat hit Primed (n = 10) Neutral (n = 10) 128 54 Pain hit 9 0 Correct rejection 143 173 Miss 43 126 False heat 22 7 False pain 15 0 Primed (n = 15) Neutral (n = 15) 63 Table 4 Total of occurrences for each possible outcome of the heat detection task during 37 °C baseline temperature trials Outcome Group Heat hit 639 306 Pain hit 127 21 Correct rejection 591 752 Miss 134 574 False heat 293 139 False pain 16 7 v2 tests were employed to examine the effect of group for each tabulated outcome with a Bonferroni correction for multiple comparisons (p < 0.05 indicates significance at the 0.05 level after correction). Two 2 · 6 v2 analyses demonstrated that the frequency of responses shown in Tables 2 and 3 both significantly deviated from chance where chance is defined as an equal distribution of responses across all the cells (v2 = 105; p < 0.05 and v2 = 84; p < 0.05, respectively). A series of 2 · 2 v2 analyses examined the tables from left to right to assess the cells describing hits (pain and heat), undetected stimuli (correct rejections and misses) and false responses (heat and pain). For the 34 °C trials these analyses yielded significant deviation from chance for undetected stimuli (v2 = 18.3; p < 0.05). Deviation from chance for hits (v2 = 3.7) and false alarms (v2 = 4.3) did not reach Bonferroni corrected criteria for significance. For the 42 °C trials the same analyses yielded significant deviation from chance for hits (v2 = 24.0; p < 0.05) and undetected stimuli (v2 = 11.5; p < 0.05), with the analysis for false alarms failing to reach a corrected threshold (v2 = 3.6). From inspection it can be inferred that the deviations in undetected stimuli for both the 34 and 42 °C trials are largely due to fewer misses in the primed group. Similarly, the deviations in hits for the 42 °C trials are largely due to increased pain hits in the primed group. 3.2. Experiment 2 Table 4 summarises the subject responses to the heat pulses generated from a 37 °C baseline for the second sample of subjects. A 2 · 6 v2 analysis demonstrated that the frequency of responses shown in Table 4 significantly deviated from chance (v2 = 544; p < 0.05). A series of 2 · 2 v2 analyses examined the table from left to right as before. These analyses yielded significant deviation from chance for hits (v2 = 20.2; p < 0.05) and undetected stimuli (v2 = 127.6; p < 0.05) but not false alarms (v2 = 0). From inspection it can be inferred that the deviations in hits are due to increased hits, especially heat hits, in the primed group and the deviations in undetected stimuli are due to fewer misses in the primed group. 3.3. Signal detection theory Fig. 1 plots the hit rate against the false alarm rate for each of the three baseline temperatures and for the neutral and primed groups and Table 5 shows the hit rate, false alarm rate, d 0 and c values for each of the relevant conditions. There is obvious and clear separation of the primed and neutral groups due to a greater hit rate and false alarm rate in the primed groups. Data from Experiment 1 were analysed using a mixed-design repeated measures ANOVA to examine the effect of baseline temperature (34 or 42 °C) and group (primed or neutral) on hit rate, false alarm rate, c and d 0 . The results of these analyses are included in Table 6. Data from Experiment 2 were analysed using oneway ANOVA to examine the effect of group (primed or neutral) on hit rate (F(1,28) = 12.6; p = 0.008), false Table 3 Total of occurrences for each possible outcome of the heat detection task during 42 °C baseline temperature trials Outcome Group Heat hit Primed (n = 10) Neutral (n = 10) 74 96 Pain hit 96 38 Correct rejection 101 147 Miss 9 46 False heat 51 27 False pain 29 6 Fig. 1. The hit rate (pain and heat hits combined) against the false alarm rate (pain and heat false alarms combined), otherwise known as the receiver operator characteristic for the task. Priming caused an obvious shift toward greater hits and false alarms for all three baseline temperatures. 64 S.S. Kirwilliam, S.W.G. Derbyshire / Pain 137 (2008) 60–65 Table 5 SDT calculations showing the hit rate, false alarm rate, c and d 0 for each group and condition SDT calculation Hit rate False alarm rate Bias (c) D-prime (d 0 ) Neutral 34 °C (SE) 0.30 0.05 1.24 1.16 (0.05) (0.01) (0.16) (0.17) Primed 34 °C (SE) 0.78 (0.06) 0.21 (0.05) 0.03 (0.15) 1.8 (0.26) Neutral 37 °C (SE) 0.48 0.17 0.50 1.21 (0.10) (0.04) (0.24) (0.33) Primed 37 °C (SE) 0.86 0.35 À0.43 1.82 (0.05) (0.06) (0.17) (0.21) Neutral 42 °C (SE) 0.72 0.21 0.11 1.65 (0.08) (0.06) (0.20) (0.24) Primed 42 °C (SE) 0.95 0.42 À0.76 1.94 (0.01) (0.06) (0.13) (0.14) c measures the distance between the criterion and the neutral point, where neither response is favoured, and is thus a measure of bias. Negative values of c signify a bias toward reporting a signal, whereas positive values signify a bias towards reporting no signal. d 0 measures the distance between the signal and the no signal means in standard deviation units. A value of 0 indicates an inability to distinguish signal from no signal, whereas larger values indicate a correspondingly greater ability to distinguish signals from noise. alarm rate (F(1,18) = 5.5; ns), c (F(1,18) = 9.9; p = 0.032) and d 0 (F(1,38) = 2.4; ns). In summary, for all baseline temperatures tested, priming caused a significant increase in hit rate with a significant shift in bias toward reporting hits but no significant change in sensitivity to the stimuli. 4. Discussion The present study examined whether priming of painrelated fear significantly altered performance on a heat detection task. Following priming it was inferred that subjects were less likely to report a stimulus as undetected and more likely to report a hit. Pain hits were more apparent for the 42 °C baseline trials and heat hits for the 37 °C trials with no significant effects at 34 °C. SDT demonstrated significant increases in hit rate and bias in the primed groups in both experiments. In addition, there was an increase in the false alarm rate in the primed group during Experiment 1 but priming did not significantly influence sensitivity in either experiment. These findings suggest a modification of responses following priming of pain-related fear with possible greater experience of pain and heat after priming even in the absence of any actual stimulus. A unique feature of this study included the use of the visual dot probe task as a method of priming painrelated fear. In contrast, prior investigations have attempted to achieve priming by simply presenting pictures that contain images loaded with emotional content (e.g. Wied and Verbaten, 2001). A limitation of this latter approach is the ineffectiveness of activating the attentional processes involved in emotion. The visual dot probe task, on the other hand, is able to tap the atten- tional resources involved with emotional processing, as indicated by the effects observed between the groups in the current study. The precise mechanism of the heightened bias observed in the primed versus neutral groups remains uncertain. One hypothesis, however, draws on the fact that individuals with chronic pain often report that pain catastrophising (i.e., anticipating the future experience of pain as horrendous or unbearable) increases subjective ratings of pain (e.g. Keefe et al., 1989; Geisser et al., 2003). Attempts to model the underlying mechanisms of this phenomenon are limited, however, Crombez et al. (2002) propose that catastrophising facilitates the priming of pain-related fear in chronic pain patients – causing hypervigilance toward painful stimuli and activating the aversive system more readily. The result is a decreased tolerance, or increased sensation, in response to painful stimulation, and activation of avoidance behaviours. It is possible that in the current study emotional priming activates a similar but temporary catastrophisation and hypervigilance in healthy subjects. From our findings, it appears that the subjects experienced increased pain and heat following priming without evidence for increased ability to discriminate signal from no-signal trials. In other words, subjects became more biased toward a report of pain or heat experience. It is possible that this increase in bias is due to other changes in attention such as an inability to disengage from threatening stimuli. The maintenance of the experience of heat and pain in the absence of any stimulus could be caused by directed attention toward the location of stimulation, an inability to disengage from the region, or both (Koster et al., 2005, 2006; Van Damme et al., 2006). It is quite likely that Table 6 Analyses of SDT parameters for Experiment 1 contrasting primed and neutral groups and the effects of temperature within groups SDT parameter Hit rate False alarm rate Bias (c) D-prime (d 0 ) Primed versus neutral F(1,18) = 28.4 (0.000) F(1,18) = 9.8 (0.048) F(1,18) = 27.1 (0.000) F(1,38) = 4.3 (ns) Temperature F(1,18) = 44.8 (0.000) F(1,18) = 38.1 (0.000) F(1,18) = 76.5 (0.000) F(1,18) = 2.5 (ns) Interaction F(1,18) = 8.5 F(1,18) = 0.5 F(1,18) = 2.3 F(1,18) = 1.1 (0.072) (ns) (ns) (ns) All thresholds corrected for multiple comparisons. S.S. Kirwilliam, S.W.G. Derbyshire / Pain 137 (2008) 60–65 65 such effects will interact with trait measurements of catastrophisation, anxiety and neuroticism as has been suggested elsewhere (Koster et al., 2006). For the current study, we omitted these measurements so as to not inadvertently prime our subjects before the heat detection task. Collecting these measures after task performance was a possibility but the trait measures might have been affected by the heat detection task and/or priming procedure. Future investigations will explore these possibilities. We have demonstrated that priming of pain-related fear can modulate sensory experience and report when receiving a potentially noxious stimulus. The results support the motivational priming hypothesis (Lang, 1995), which predicts that negative emotions increase negative sensory experiences via activation of the aversive system. In addition, the findings provide indirect evidence that biases in attention of potentially noxious stimuli facilitate a bias toward heat and pain detection in primed individuals. The visual dot probe task appeared to be an effective method of priming painrelated fear. The visual dot probe and heat detection task described here may have application in a clinical setting, such as an objective assessment of hypervigilance toward pain among individuals with chronic pain and other somatoform disorders. Acknowledgements We thank Lucie Clifford for additional data collection during Experiment 2 and Tracy Warbrick, Andrew Welchman and Lotte Meteyard for considerable statistical advice. Many thanks to the two anonymous reviewers whose efforts to improve our report were insightful, valuable and beyond the call of duty. References Bantick SJ, Wise RG, Ploghaus A, Clare S, Smith SM, Tracey I. Imaging how attention modulates pain in humans using functional MRI. Brain 2002;125:310–9. Bradley BP, Mogg K, White J, Groom C, de Bono J. Attentional bias for emotional faces in generalized anxiety disorder. Br J Clin Psychol 1999;38:267–78. Crombez G, Eccleston C, Van den Broeck A, Van Houdenhove B, Goubert L. The effects of catastrophic thinking about pain on attentional interference by pain: no mediation of negative affectivity in healthy volunteers and in patients with low back pain. Pain Res Manag 2002;7:31–9. Eccleston C, Crombez G. Pain demands attention: a cognitive-affective model of the interruptive function of pain. Psychol Bull 1999;125:356–66. Fulbright RK, Troche CJ, Skudlarski P, Gore JC, Wexler BE. Functional MR imaging of regional brain activation associated with the affective experience of pain. AJR Am J Roentgenol 2001;77:1205–10. Geisser ME, Casey KL, Brucksch CB, Ribbens CM, Appleton BB, Crofford LJ. Perception of noxious and innocuous heat stimulation among healthy women and women with fibromyalgia: association with mood, somatic focus, and catastrophizing. Pain 2003;102:243–50. Good M, Stanton-Hicks M, Grass JA, Cranston G, Anderson GC, Choi C. Relief of postoperative pain with jaw relaxation, music and their combination. Pain 1999;81:163–72. Janssen SA, Arntz A, Bouts S. Anxiety and pain: epinephrine-induced hyperalgesia and attentional influences. Pain 1998;76:309–16. Keefe FJ, Brown GK, Wallston KA, Caldwell DS. Coping with rheumatoid arthritis pain: catastrophizing as a maladaptive strategy. Pain 1989;37:51–6. Kenntner-Mabiala R, Pauli P. Affective modulation of brain potentials to painful and nonpainful stimuli. Psychophysiology 2005;42:559–67. Koster EH, Verschuere B, Crombez G, Van Damme S. Time-course of attention for threatening pictures in high and low trait anxiety. Behav Res Ther 2005;43:1087–98. Koster EH, Crombez G, Verschuere B, Van Damme S, Wiersema JR. Components of attentional bias to threat in high trait anxiety: Facilitated engagement, impaired disengagement, and attentional avoidance. Behav Res Ther 2006;44:1757–71. Koyama Y, Koyama T, Kroncke AP, Coghill RC. Effects of stimulus duration on heat induced pain: the relationship between real-time and post-stimulus pain ratings. Pain 2004;107:256–66. Lang P. The emotion probe: studies of motivation and attention. Am Psychol 1995;50:372–85. Larsson J, Landy MS, Heeger DJ. Orientation-selective adaptation to first- and second-order patterns in human visual cortex. J Neurophysiol 2006;95:862–81. Meagher MW, Arnau RC, Rhudy JL. Pain and emotion: effects of affective picture modulation. Psychosom Med 2001;63:79–90. Miron D, Duncan GH, Bushnell MC. Effects of attention on the intensity and unpleastness of thermal pain. Pain 1989;39:345–52. Peyron R, Garcia-Larrea L, Gregoire MC, Costes N, Convers P, Lavenne F, et al. Haemodynamic brain responses to acute pain in humans: sensory and attentional networks. Brain 1999;9:1765–80. Stanislaw H, Todorov N. Calculation of signal detection theory measures. Behav Res Meth: Instruments & Computers 1999;31:137–49. Van Damme S, Crombez G, Hermans D, Koster EH, Eccleston C. The role of extinction and reinstatement in attentional bias to threat: a conditioning approach. Behav Res Ther 2006;44:1555–63. Villemure C, Bushnell MC. Cognitive modulation of pain: how do attention and emotion influence pain processing? Pain 2002;95:195–9. Wicch K, Seymour B, Kalisch R, Stephan KE, Koltzenburg M, Driver J, et al. Modulation of pain processing in hyperalgesia by cognitive demand. Neuroimage 2005;27:59–69. Wied M, Verbaten MN. Affective picture processing, attention, and pain tolerance. Pain 2001;90:163–72.