Strengthened endogenous pain modulation by merging CPM with distraction: also applicable for fibromyalgia patients?
Pain is the primary reason for people to visit a doctor (Mantyselka et al., 2001). Pain is defined as “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage” (Merskey, 1986, p. 210). It is functional as it warns us of potential damage to our body and often disappears once the damaged tissue is healed (Morrison & Bennett, 2012). However, when pain lasts for more than three to six months and fails to improve over time, it becomes chronic and problematic. Some chronic pain syndromes have an identifiable cause (e.g., rheumatoid arthritis), while others do not (Deyo, 1991). An example of the latter is the chronic pain syndrome fibromyalgia (FM), which is a syndrome that is characterised by pain, stiffness, and fatigue (Clauw, 2014; Croft, Schollum, & Silman, 1994). FM is further accompanied by symptoms such as sleep disturbances, depression, and anxiety. FM affects about 2% to 8% of the population and seems to have a greater prevalence in women (3 to 5%) than in men (0.5 to 1.6%; Clauw, 2014; White, Speechley, Harth, & Ostbye, 1999; Winfield, 2000). The American College of Rheumatology (ACR) were first to develop criteria for diagnosing FM: Pain by pressure on eleven out of eighteen tender points and widespread pain in every quadrant of the body for at least three months (Goldenberg, Burckhardt, & Crofford, 2004; Wolfe et al., 2010). Later, the Widespread Pain Index replaced the tender points, and the Symptom Severity Scale was included (Wolfe et al., 2010). To be diagnosed with FM, all symptoms must be verified and all other possible causes must be ruled out (Clauw, 2014; Buskila & Zarzi-Puttini, 2006). Although FM is widely researched, the pathophysiology and etiology still is poorly understood, making treatment of FM challenging for clinicians. There are some effective treatments available for patients (e.g., Cognitive Behavioural Therapy and Psycho-education; Goldenberg et al., 2004), though no intervention seems to be effective for all patients (Clauw, 2014), and there is no standard treatment (Okifuji & Hare, 2013). It is therefore important to learn more about the underlying pain modulation systems and how psychological factors are involved.
It has been suggested that a deficit in endogenous pain modulation systems may contribute to chronic pain syndromes such as FM (Hendrikson & Mense, 1994; Lewis, Rice, & McNair, 2012; Yunus et al., 1992; Van Wijk & Veldhuijzen, 2010). In healthy controls, endogenous pain modulation systems possess the capacity to inhibit or enhance the nociceptive input, and are collectively termed endogenous analgesia (Yarnitsky et al., 2010). One of the measures for endogenous analgesic mechanisms is termed as ‘diffuse noxious inhibitory controls’ (DNIC), which is extensively studied in animal models over the years (Cadden, Villanueva, Chitour, & Le Bars, 1983; Le Bars, Dickenson, & Besson, 1979), and is used to find dysfunctions in endogenous analgesia (Pud, Granovsky, & Yarnitsky, 2009). DNIC is commonly known as the ‘pain-inhibits-pain’ phenomenon (i.e., pain in one region of the body can inhibit pain in another; Yarnitsky et al., 2010). Functionally seen, DNIC is a mechanism that acts like a barrier and keeps pain bearable and regional (Pielsticker, Haag, Zaudig, & Lautenbacher, 2005) as it allows pain-signalling neurons in the trigeminal nuclei and the spinal dorsal horn to be inhibited by a noxious stimulus applied to a body area far remote from the pain-signalling neurons (Le Bars et al., 1979). In a standard DNIC experiment, one body region receives a painful stimulus (the conditioned stimulus – CS), while the pain evoked by another stimulus administered to another region (the test stimulus – TS), is measured. The measured reduction in pain scores of the latter is indicative of DNIC (Defrin, Tsedek, Lugasi, Moriles, & Urca, 2010). Similarly, this ‘pain-inhibits-pain’ phenomenon is extensively studied in humans by using electrophysiological and psychophysical tools (e.g., Granot, et al., 2008; Pud et al., 2009; Pud, Sprecher, & Yarnitsky, 2005). Recently, Yarnitsky et al. (2010) introduced the term ‘conditioned pain modulation’ (CPM) to describe the psychophysical paradigm of testing DNIC in humans. Therefore, the term CPM will be used in this article. Studies on CPM have found reductions of sensitivity to periodic pain when a pain stimulus was applied (Lautenbacher, Roscher, & Strian, 2002). Consequently, a deficit in CPM could be the cause of the presence of hyperalgesia (increased sensitivity to pain) and allodynia (increased response of touch) among some FM patients (Staud, 2009). Some studies have observed dysfunctions of CPM effects in FM patients. For example, Kosek and Hansson (1997) and Lautenbacher and Rollman (1997) show reduced experimental pain sensitivity in healthy controls, but not in FM patients, after applying tonic conditioning nociceptive stimulation. Also, exercise during stimulation seemed to reduce TS pain scores in healthy subjects, but not in FM patients (Vierck et al., 2001). It remains however the question whether a deficient CPM effect is indicative of developing chronic pain (Yarnitsky et al., 2008), or whether CPM slowly wears out when pain remains as it cannot be maintained for a long time as suggested by Van Wijk and Veldhuijzen (2010).
As described, CPM is an effective way to reduce perceived pain of stimuli. Another effective way to influence the perception of pain is to draw the attention away from the pain source, resulting in a decrease of perceived pain intensity (e.g., Moont, Crispe, Lev, Pud, & Yarnitsky, 2012; Villemure & Bushnell, 2002). Subsequently, attending to a painful stimulus can increase the perceived pain intensity. Research shows that distracting methods such as listening to music can significantly reduce perceived post-operative pain (Good et al., 1999; Good et al., 2000), while patients who are required to attend to pain report more intense post-operative pain (Miron, Duncan, & Bushnell, 1989). Moreover, the findings of James and Hardardottir (2002) examined the relationship between pain and attention by asking participants to place their lower arm in freezing cold water while concentrating to either a computer based task or the pain sensations. Participants who focused on the pain sensations pulled their arm significantly earlier out of the cold water than those who focused on the distraction task. These findings are also supported by the studies of Bantick et al. (2000), Veldhuijzen et al. (2006), and Valet et al., (2004). They suggest that pain perception can be affected by task performance. In their studies, participants perceived pain as less intense when they were performing demanding attention tasks, because their attention was distracted away from the pain. As distraction also inhibits painful stimuli, it raises the question how attention determines CPM and whether CPM is partly or totally due to attention manipulation.
It has been suggested that CPM effects might be largely due to distraction: When attention is drawn away from the TS it will presumably result in lower perceived pain scores. However, after examining this hypothesis, multiple studies suggest that CPM and distraction work independently and that CPM is minimally related to attentional processes (Van Wijk & Veldhuijzen, 2010). Kakigi (1994) used CO2 laser stimulation and painful somatosensory evoked potentials to induce CPM effects in healthy volunteers. To measure the effect of attention on the CS, subjects were asked to point their attention at the knee where the CO2 laser stimulation was applied. The results of Kakigi (1994) show that the effect of CPM does not solely rely on distraction by the CS, because CPM effects were still found when attention was intentionally directed to the TS when the CS was administrated. CPM effects also remained unchanged when Edwards, Fillingim, and Ness (2003) controlled for subjective distraction scores by the CS in their study. Finally, in the study of Lautenbacher, Prager, and Rollman (2007), healthy subjects received tonic heat stimuli to the skin on the thigh delivered by the use of a Peltier thermode (Phywe Systeme, Goettingen, Germany). Painful electrical stimuli were delivered to the skin on the volar forearm to generate the CPM effect. A visual distraction task was included to measure if CPM is (partly) moderated by attention. In the distraction condition, signal lights were activated and the subjects were required to provide a pain rating before they reported the number of lights they counted. Their findings showed that CPM effects were not influenced by distraction, though the distraction task produced a small but significant decline in pain ratings. They concluded that CPM and distraction might work as two separate pain inhibition mechanisms. Overall, these results suggest that CPM effects are relatively insensitive to attentional processes and that CPM works as an independent pain inhibition mechanism. However, these results may be biased due to the overt nature of the instructions, and there was no reliable objective way of measuring participants’ compliance during the distraction stimuli. Also, Lautenbacher et al. (2007) did not analyse changes in test pain scores, which is normally included in ‘classic’ CPM conceptions.
In an attempt to resolve these issues, Moont, Pud, Sprecher, Shartvit, & Yarnitsky (2010) used a continuous cognitive distraction task, in which active involvement of attentional resources was required. This task was furthermore presented as a computer game, not as an overt way of drawing attention away from painful stimuli. Moreover, Moont et al. (2010) measured the number of correct responses instead of relying on a subject’s compliance. They used heat stimuli (TS), a conditioning stimulus (hand in hot water bath) and a visual search task as the distracter in their study, to examine if CPM and attentional processes work as two separate inhibitory mechanisms. Moreover, they hypothesised that applying these two pain modulation paradigms simultaneously would have an additive inhibitory effect on pain stimuli. Pain intensity ratings were measured in the CPM, distraction, and combined conditions. During the distraction and combined conditions, participants were instructed to complete this visual search task. In this task participants were shown different groups of coloured shapes every 1.5 seconds, in three different levels of task load. The subjects had to count the appropriate shapes mentally and had to report the total verbally when requested. It was suggested that increasing cognitive attention load was related to the increasing distraction task load. The findings of Moont et al. (2010) showed a decrease of pain scores in both the CPM condition and the distraction condition compared to the baseline TS, suggesting effective pain inhibition in both conditions. This seems to confirm the hypothesis that pain modulation via CPM may not be due to attention modulation and could a physiological mechanism on its own. Moreover they found small but significant additive pain inhibition when CPM and distraction were combined, compared to the CPM condition alone. This suggests that CPM and distraction may have an additive effect on reducing both intensity and unpleasantness pain ratings of the test pain, when they are applied simultaneously. The study of Moont et al. (2010) was the first in which the effect of continuous cognitive distraction on CPM was specifically measured. However, this type of experiment, where the conditions of attention and CPM were assessed separately, has never been performed on FM patients. Nevertheless, attention seems to be relevant in pain modulation in FM patients. Staud et al. (2003) found a small but significant additive effect on CPM when participants directed their attention to the TS and CS. Though, this was only the case in female FM patients and not in healthy controls. CPM and distraction conditions were not assessed separately, so no valuable assumptions can be drawn from this result. Nevertheless, differences in attentional function in FM patients and healthy controls may be of influence in the ability to modulate endogenous pain. However, studies on attentional performance in FM patients show conflicting results regarding the cause of reduced attentional performance in FM patients.
Some researchers claim to have found evidence of a decreased performance on tasks examining the attentional function in FM (e.g., Dick, Verrier, Harker, & Rashiq, 2008; Glass, 2009). In contrast to this evidence, others have failed to show this decrease when controlling for sleep disruptions, depression, anxiety, pain, and fatigue (Suhr, 2003; Wallace, 1997). These confounding factors seem to be common features in FM patients and are associated with cognitive disruption (e.g., Amutio, Franco, Pérez-Fuentes, Gázquez, & Mercader, 2015; Torta, Pennazio, & Ieraci, 2014). Moreover, Oosterman, Derksen, Van Wijk, Kessels, and Veldhuijzen (2012) examined the effects of chronic pain on both attentional and executive functions, as attention is difficult to segregate from executive functions at a conceptual level, because of the strong overlap. They also took psychomotor speed and possible confounders such as pain, catastrophizing, and depression into account. Oosterman et al. (2012) used the Stroop Colour-Word Task to measure cognitive inhibition – an external validated test for cognitive inhibition (Breton et al., 2011). This task seems to be a judicious choice for assessing cognitive inhibition in chronic pain syndromes as it is presumably closely related to pain sensitivity and is suggested to be related to altered pain modulation (Oosterman, Dijkerman, Kessels, & Scherder, 2010). The results of Oosterman et al. (2012) suggest that executive and attentional functions appear to be largely intact in chronic pain patients, except for the significant decline in the ability to sustain attention. This decline may indicate a decreased attentional function in chronic pain patients and may suggest that these observed attentional deficits may be related to chronic pain (Dick, Eccleston, & Crombez, 2002). The latter corresponds to the findings of Eccleston (1995), Eccleston et al. (1997), and Grisart and Plaghki (1999). Their findings demonstrate that chronic pain acts as an equivalent of controlled processing and therefore consumes a portion of the limited attentional resources in chronic pain patients. However, there are a lot of conflicting findings regarding attentional functions in chronic pain patients, and the precise pattern of performance on attention task in FM patients remains unclear. Moreover, it raises the question whether chronic pain effects, or a deficit in attentional function, play a role in the decreased endogenous pain modulation in FM patients.
In summary, although CPM is extensively studied, it still requires further investigation (Defrin et al., 2010), because it is still unknown how cognitive factors (e.g., attention/distraction) determine CPM. Few studies examined attentional function in CPM and more research is needed to examine the potential effect of attentional manipulation in CPM on endogenous pain modulation. In the study of Moont et al. (2010) a visual search task was used as the distraction task. As it has been suggested that altered pain modulation and pain sensitivity are possibly related to cognitive inhibition and not to other forms of executive attention (Oosterman et al., 2010), the present study therefore used the Stroop Colour-Word Task as the distraction task. Also, the present study was the first to examine the effects of CPM, distraction, and CPM and distraction combined in FM patients. As research on attentional function in FM patients show conflicting results, it remains unclear whether a deficit in attentional function is the cause of reduced pain inhibition during distraction in FM, or the consequence of chronic pain such as depression and anxiety. Attentional and executive function domains show strong overlap. The present study therefore included an extensive assessment of several attentional function domains. More clarity may lead to improvements in the healthcare of FM patients, as it may clarify where to focus on in pain management.
The present study therefore aimed to provide more insights into CPM and to demonstrate what role attention has in this endogenous pain modulation in healthy controls and FM patients. The primary research question stated: What is the role of distraction on the CPM effect? It was expected that CPM and distraction work as two independent pain inhibition mechanisms as demonstrated in the study of Moont et al. (2010). A small but additive inhibitory effect on pain stimuli was expected when both CPM and distraction are applied simultaneously. The secondary research question stated: Is there a difference between FM patients and healthy controls for the pain inhibitory effect if CPM and distraction are combined? By comparing the FM patients with the healthy controls, healthy controls were expected to have a stronger pain inhibitory effect than FM patients when CPM and distraction are applied simultaneously. A weaker pain inhibitory effect in FM patients was expected because of a possible CPM deficit (Kosek & Hansson, 1997; Lautenbacher & Rollman, 1997; Vierck et al., 2001) and a possible decreased ability to sustain attention to the distraction task (Oosterman et al., 2012), which presumably reduces the additive inhibitory effect on CPM. The tertiary research question stated: Can this difference in endogenous pain inhibition be potentially explained by differences in attentional function between FM patients and healthy controls? It was expected that a difference between the groups for this pain inhibitory effect is most likely due to the result of a general deficit in attentional function in FM patients (Oosterman et al., 2012; Dick et al., 2002).
To answer these questions, the experimental protocol of Moont et al. (2010) was used. Solely healthy controls were examined to answer the first question. This led to a clearer picture of what role attention plays in endogenous pain modulation, and if distraction acts additively to the CPM effect. The second question was examined by comparing FM patients with healthy controls. CPM, distraction, and combined effects (CPM with a distraction task) were examined in both groups. The Stroop Colour-Word Task was used to assess cognitive inhibition and served as the distraction task. The computerised neurocognitive Attention Network Test (ANT) was used to measure the three components of attention: Orienting, alerting and executive function (Fan, McCandliss, Sommer, Raz, and Posner, 2002; Fan et al., 2009). To determine if possible differences in attentional performance are dependent on the attentional component of executive functioning, the Key-Search subtest of the Behavioural Assessment of the Dysexecutive Syndrome (BADS) was used. This subtest measures executive functioning with little dependence on attentional processes (Wilson, Evans, Alderman, Burgess, & Emslie, 1997). Possible confounders were controlled for to examine if differences between the groups are caused by a deficit in attentional function, or because of chronic pain effects.
Method
Participants
A total of 73 participants were recruited for the present study; 20 female and 13 male patients diagnosed with FM, and 20 healthy female and 20 healthy male controls. Less male patients were recruited since most patients are female (e.g., Clauw, 2014), and male patients met the exclusion criteria more often. Staff members of the Rheumatology and Clinical Immunology department informed FM patients about the study, and healthy controls were recruited through advertisements in public access rooms in the UMC Utrecht, around campus, or through announcements in local newspapers. Healthy subjects were matched on educational level and age to FM subjects. Further, to make sure the participants met the inclusion criteria and not the exclusion criteria, all participants had to fill out a biographic questionnaire. The inclusion criteria were: (a) All FM patients have received the diagnosis FM according to the American College of Rheumatology classification criteria (Wolfe et al., 1990) by a rheumatologist; (b) All subjects speak Dutch fluently, (c) All subjects are adults (> 18 years old); (d) Healthy controls were matched to patients on educational level and age, and are pain-free as determined by a general health questionnaire. Subjects were excluded from participating when: (a) They were unable to give informed consent; (b) They had neurobiological or psychiatric conditions besides FM; (c) They used sedative psychotropic or analgesic drugs; (d) They had serious injury on the body regions to be tested; (e) They participated in another research protocol that could influence or interfere with the outcomes of this study.
The present study is a randomized repeated measures study. To avoid expectancy effects, all subjects were blinded with regard to temperatures used and to the hypotheses of this study. The order of the experimental blocks was semi-randomised across participants. To prevent learning effects to affect the comparison between task performance over groups, the experimental conditions were randomised during the experimental runs. Hereby, an equal proportion of subjects starting with the Combined Block or Distraction Block was guaranteed.
Measures
Pain Ratings. A Numerical Rating Scale (NRS) was used for subjects to rate their pain immediately after stimulation. A copy of the NRS was constantly observable for the subjects during the tests. The scale was composed of integers in increments of 5 with a verbal description associated with increments of 10: 0 = no pain, 10 = very weak pain, 20 = weak pain, 30 = weak to moderate pain, 40 = moderate pain, 50 = moderate to high pain, 60 = high pain, 70 = high to very high pain, 80 = very high pain, 90 = nearly intolerable pain, and 100 = intolerable pain.
Training and sensitivity. The experimental protocol of Moont et al. (2010) was used for the experimental runs. The TS (primary painful stimulus) was applied to the left volar forearm, and the CS (secondary painful stimulus) was applied to the right hand, simultaneously with the TS. Both stimuli were adjusted to elicit a 50 (moderate pain) on a numerical rating scale (NRS) from 0 (no pain) to 100 (intolerable pain) immediately after stimulation. These temperatures were determined at the beginning of the session. Also, subjects were trained in rating painful stimuli following stimulation. Moreover, all subjects were blinded with regard to the temperatures used in this study.
The temperatures of the TS was delivered by a Peltier-type computerised thermal stimulator (Cheps, Pathway, Medoc, Israel) that generates heat stimuli using a flat probe surface contacting a cutaneous area of 3×3 cm. It can heat up rapidly to a maximum rate of 70 °C/s and has safety margins that are built-in. From a baseline temperature of 32°C, the intensity of the temperature was increased up to a never exceeding maximum of 53°C, depending on the individual target temperature. First, to determine the heat pain thresholds (HTP) and to reduce surprise effects, warm temperatures of the TS were continuously increased until the participant pressed the button, indicating the first painful hot sensation (thus, a pain rated of 50 on the 0 to 100 NRS). Warm temperatures for both CS and TS were used to exclude potential distraction effects (Moont et al., 2010). This procedure was repeated three times and the average was used to determine the HTP.
Hereafter, the pain intensity of the TS was determined by using short series of descending and ascending test stimuli between 35 °C and 53°C. Again, subjects had to identify their pain intensity score of 50 on the 0-100 NRS. This part took approximately 10 minutes to complete. Further, to measure the CS the participants had to put their right hand up to the wrist into a warm water bath. They had to indicate the first pain sensation and were asked to indicate this sensation after 30 seconds of immersion. They were instructed to remove their hand from the bath if the pain was intolerable. The temperature of the water was increased or decreased depending on the pain rating after 30 seconds. This was to target a NRS of 50. This test was repeated two times after adjusting the water to the right painful temperature. The duration of the training and sensitivity part took about 15 minutes.
Attention Network Test (ANT). The Attentional Network Theory proposes three independent cognitive concepts: Alerting (i.e., the capacity to achieve and sustain a vigilant state), orienting (i.e., focused identification and selection of sensory information), and executive functioning (i.e., the capacity to monitor and resolve conflicts between competing processes; Fan, McCandliss, Sommer, Raz, & Posner, 2002). The ANT provides a measure of the efficiency of these attentional networks. It is a combination of a flanker task (Eriksen & Eriksen, 1974) and a cued reaction time task (Posner, 1980). In the computerised version of the ANT, participants were shown consecutively presented stimuli on a computer screen using E-Prime software (Psychology Software Tools, Pittsburgh, PA). Subjects had to perform a short practice task on the PC before carrying out the actual test.
The stimuli presented consisted of a row of five horizontal black lines, with arrows pointing right or left. The target participants had to focus on a right or left arrow presented at the centre of the screen. Depending on the direction of the arrow, participants had to respond on one of two buttons on the keyboard. Incongruent (executive), congruent, or neutral distracters flank the target arrow. Before the target was presented, participants were either given a centre temporal cue (alerting), a spatial cue (orienting), or no cue (control). In the incongruent condition, five arrows are flanking the target point in opposite direction from the target arrow, while all five arrows point in the same direction in the congruent condition. Subjects had to indicate as fast and accurate as possible to which side the target arrow was pointing. Each trial consisted of five events and began with a fixation cross in the middle of the screen. First, a target display appeared for a variable duration of 400 to 1600 ms. Next, a warning cue was presented for 100 ms. After the warning cue there was a fixation period for 400 ms. Hereafter the target and flankers simultaneously appeared. The flankers and target were presented until the subject made a response (up to a maximum of 1700 ms), and then disappeared. A post target fixation period appeared after the response, which was based on the duration of the first fixation and reaction time (3500 ms minus the duration of the first fixation minus the reaction time). Hereafter the next trial began. Each trial lasted 4000 ms. Scores were measured by comparison of the mean reaction time between one condition and the appropriate reference condition. The alerting effect was calculated by subtracting the mean reaction time of the double-cue conditions from the mean reaction time of the no-cue condition. The orienting effect was calculated by subtracting the mean reaction time of the spatial cue conditions from the mean reaction time of the centre cue. The executive control effect was measured by subtracting the mean reaction time of all congruent flanking conditions (summed across cue types), from the mean reaction time of incongruent flanking conditions (Fan et al., 2002).
Stroop Colour-Word Task. The Stroop Colour-Word Task measures cognitive inhibition, a specific form of executive attentional performance (Stroop, 1935). In order to give the effortful correct response in this task, the participants had to inhibit their automatic response. During the experimental session, three cards were presented to the participants. Each card contained 100 words (10 rows with 10 colour names) with an increasing task load. The participants were instructed to respond to each stimulus as fast as possible without stopping. In the first condition subjects had to read colour words written in black ink (red, blue, yellow, and green). This card was used to prime the subjects with reading word aloud quickly. In the second (nonword) condition, subjects had to name the colours of rectangles printed in colour ink. The third condition specifically addresses inhibition. Subjects had to name the colour of colour words, but the colours were incongruent with the words. The latter causes the interference effect. There was no time limit to complete the conditions. The numbers of errors made, the amount of self-corrections, and the total time per card were measured. To establish diminished cognitive inhibition performance in FM, a disproportional decline in performance should be present in the third condition. An interference measure was calculated by using the classical method: Time measured for completing the second condition was subtracted from the time measured to complete the third condition (Hammes, 1971; Breton et al., 2011). The Stroop Colour-Word Task was completed before the experimental runs to measure differences in performance between groups. Furthermore, the Stroop Colour-Word Task was used as distraction task: The third condition was repeated twice for 30 seconds during CPM testing.
Behavioral Assessment of the Dysexecutive Syndrome (BADS). The BADS includes six tests and two questionnaires. In the present study, the Key-Search subtest of the BADS (Wilson, et al., 1997) was used to determine if differences in attentional performance are dependent on a general deficit in attentional function or an attentional component. This subtest measures executive functioning with little dependence on attentional processes. The test consisted of a 100 mm square on the middle of an A4 sized sheet with a small black dot 50 mm below the square. The subjects were asked to draw the path they would walk to find their key they lost within a large imaginary field. The starting point was the black dot. The Key-Search subtest can be used to assess a person’s ability to create a plan of action and how to solve a problem. This requires memory and scanning abilities. Scores range from 0 to 16 and are based on a number of criteria, including whether the rater believes the used strategy to be quick, efficient, effective, and systematic (Katz, Tadmor, Felzen, & Hartman-Maeir, 2007).