Putting the brain to work in cognitive-behavioral therapy for chronic pain

Abstract

Healthcare practitioners may not be familiar with biopsychosocial therapies such as cognitive-behavioral therapy (CBT), and may view psychosocial treatments as only appropriate for those with mental illness. Further, practitioners may not know that treatments based on a biopsychosocial model have a biological basis. Treating patients with pain as if all pain is related to ongoing tissue damage that must be located and repaired (a biomedical model) has led to the false impression that hurt = harm and that “real” pain is detectable with a biomedical marker and correctable with the right biomedical intervention. Many of our patients do not know that the brain is critically important to the perception of pain and in pain self-management. In this review, I describe recent research using a simplified gate-control schematic as a treatment rationale for biopsychosocial pain education and CBT with low-income, multiply disadvantaged patients. Using the cognitive model of pain (simplified as “Think → Feel → Act”), I provide examples of how psychosocial health care providers train patients in skills they can independently use to work with their thoughts to motivate and implement pain self-management behaviors. Moving into a brief examination of treatment mechanism (mediation, moderation), I will conclude with the argument that effective psychosocial interventions share more commonalities than sometimes acknowledged, and that these treatments can be used as part of interdisciplinary treatment to help patients self-manage their pain.

1. Introduction

Biomedical treatments for chronic pain have not been uniformly effective and sometimes have caused harm [38,40,76]. As a result, international guidelines have proliferated for treating chronic pain using nonpharmaceutical, noninvasive, biopsychosocial therapies [3,24,37,68,94]. In current practice, however, clinics delivering unimodal biomedical interventions are flourishing (and such treatments are increasingly provided even in developing countries) [5,7]. Meanwhile, interdisciplinary pain treatment programs, despite their superiority over unimodal treatments [25,39,87], have steadily declined. Based on U.S. Medicare reimbursement data available over 20 years ago, when laminectomy surgery costs were approximately $30,000 USD, a standardized efficacious 10-session group cognitive behavioral treatment (CBT) approach cost 168 times less than spinal surgery [9]. More recent data reports skyrocketing costs for spinal surgeries, some of which bill out at over $400,000 USD for the surgery alone [38]. Meanwhile, biopsychosocial pain treatments continue to be poorly reimbursed (if at all), and are not widely available in the U.S. or internationally.

Many healthcare practitioners are not familiar with biopsychosocial therapies such as CBT, and may view psychosocial treatments as only appropriate for those with mental illness. Moreover, some practitioners may hold the erroneous belief that chronic pain is an expression of mental hurt and generated psychologically. Further, practitioners may not know that treatments based on a biopsychosocial model have a biological outcome. However, a wealth of evidence supports that biopsychosocial treatments are conceptually and empirically linked to our current understanding of pain neurophysiology. For example, there is growing evidence that CBT can reverse some of the neuronal changes associated with chronic pain, including post-treatment increases in the prefrontal and posterior parietal cortex (areas thought to be associated with pain appraisal), and decreases in somatosensory cortex (perhaps reflecting reductions in the perception of pain severity) [72]. A randomized controlled trial of CBT compared to pain education reported post-CBT decreases in connectivity between the limbic system (emotions) and other areas of the brain, and enhanced connectivity between prefrontal cortices (cognition) and other regions of the brain [74].

Treatments using the biopsychosocial model do not ignore the biological, nor do they advocate refraining from appropriately assessing and treating ongoing tissue damage. However, treating patients with pain as if all pain is related to ongoing tissue damage that must be located and repaired (a biomedical model) has led to the false impression that hurt = harm and that “real” pain is detectable with a biomedical marker and correctable with the right biomedical intervention. The net result is that, at least in the U.S., healthcare costs are out of control, there is an opioid epidemic, and patients have unrealistic expectations of being pain-free. Many of our patients do not know that they can harness the power of their thoughts, emotions, and behaviors to help themselves manage their chronic pain, thus increasing function and quality of life despite pain.

In this review, which is written as part of the 2020 IASP Ronald Melzack Lecture Award, I revisit Melzack and Wall’s gate control model of pain [54] as a biopsychosocial construct, and explore its more recent clinical applications in specific kinds of pain education [56,79,83] and cognitive-behavioral therapy - CBT [79,83]. Although it is acknowledged that we have advanced our understanding beyond the original gate control theory to more fully appreciate the role of the brain in pain, I will discuss the applicability of using this very basic framework for patients. I describe recent research using a simplified gate-control schematic as a treatment rationale for pain education and CBT with low-income, multiply disadvantaged patients [83], using the gate control model as a launching point to discuss the cognitive aspects of CBT (and pain education). Targeting a cognitive model of pain (thoughts influence emotions and behavior), I provide examples of how psychosocial health care providers work alongside patients to train them in skills to work with their thoughts, to motivate patients, and to help them implement pain self-management behaviors. Moving into a brief examination of treatment mechanism (mediation, moderation), I will conclude with the argument that effective psychosocial interventions share more commonalities than sometimes acknowledged, and that these treatments can be used to help patients self-manage their pain, and perhaps to change healthcare cultures.

2. Gate Control Model Briefly Revisited

More than 50 years ago, Ronald Melzack and Patrick Wall [54] proposed that a neural gate in the spinal cord could modulate pain signals coming from the periphery and going to the brain. Today, this is common knowledge in the pain research world, but in 1965, we were still operating under the sixteenth Century Cartesian notion that nociceptors in the periphery had a direct line to the brain (via spinal cord). Thus, pain signals, activated by noxious stimuli, rang a bell in the pain center of the brain. Presumably, the more nociceptors that were activated (i.e., the stronger the stimulus), the louder the bell rang. It is now clear that there is not a one-to-one correspondence between tissue damage and pain, and that one’s experience of pain does not necessarily indicate ongoing tissue damage. However, there is still an alarming amount of adherence to a strict biomedical model of pain within a variety of cultures, and in many healthcare systems around the world. What was revolutionary about the gate control model was the idea that the brain could monitor and modulate the activity of the spinal cord transmission neurons. Specifically implicated were brain-mediated cognitive and affective factors as part of the neural process of pain perception, and part of the descending “central control” of the spinal cord gating system.

Although it is now established that neural mechanisms are more complex than that originally proposed by Melzack and Wall [52–54], the general concepts remain; that a gating mechanism in the spinal cord excites or inhibits information flow to the brain, and that there are descending central control mechanisms that increase or decrease the sensitivity of the nociceptive system via this gating mechanism -- possibly, via multiple gates and multiple systems [8,86]. The perception of pain is influenced by signals received from the periphery, but driven by what the central nervous system does with those signals.

3. Gate Control Model Employed Clinically

Educating patients about how pain is processed in the brain (including elements of the gate control/neuromatrix model) is considered to be of increasing importance as reflected in international pain management guidelines as well as its inclusion in many current biopsychosocial treatment approaches. Lorimer Moseley and colleagues have promoted pain neurophysiology education (PNE) for patients with chronic pain [11,51,55,57,58,60,73], and there is a growing literature examining its efficacy. Systematic reviews and recent clinical trials report differing levels of support for the “neuroeducation” approach, particularly as a unimodal treatment, but there is strong support for the rationale of explaining pain processing in the brain when working with individuals in pain [15,46,85,95,96]. Of note, one such review reported that multi-session cognitive-behavioral therapy (CBT) preceded by pain neurophysiology education produced better longer-term outcomes compared to neurophysiology education alone [15]. Another clinical trial reports that hypnosis enhances the effect of PNE in patients with low back pain and a very recent paper suggests combining PNE with motivational interviewing techniques may bolster its effects [61,70]. Thus, pain neurophysiology education may not be enough.

In past and current psychosocial research, pain education alone has sometimes been used as an attention control condition [31,88]. Although there are multiple approaches to a pain education intervention, and reported efficacy is variable [2,63], there is likely an important distinction between pathophysiology education (which provides a biomedical explanation of disc pathology, for example) and a biopsychosocial approach. The latter typically includes pain self-management education, often provides some measure of neurophysiology education, and usually emphasizes the importance of cognitions and emotions [29,34,49,79,82–84,90]. Psychosocial pain education could be considered a therapeutic strategy if it is formulated to enhance patient understanding of their condition, help them reconceptualize pain and pain management as multi-dimensional rather than solely determined by tissue pathology, and provide a rationale for engaging in pain self-management. A key point is that for pain education to be effective, it must provide the rationale (and likely, the motivation) to engage in known pain self-management behaviors. In either case, behaviors must change for pain self-management to be effective. For example, if patients learn about the importance of maintaining physical activity despite chronic pain, physical activity must actually be implemented to effectively reduce pain-related interference in daily activities.

Our clinical research team has incorporated into our comprehensive treatment approach a simplified discussion of pain as a perception that is filtered by the brain, emphasizing the capacity of the brain to “open” or “close” a pain gate to process more or fewer pain signals. See Figure 1 for the illustration we use to facilitate this discussion. This is an adaptation of more complex pain neurophysiology education approaches [11,12] and it serves as our treatment rationale for psychosocial pain management, specifically, cognitive-behavioral therapy [32,79]. We emphasize the following specific points: First, contrary to what is typically understood, pain is not a sensation, but a perceptual experience in the brain that involves many filters. Second, the brain processes the sensation of pain by filtering so-called pain signals through multiple sites involved in memory, emotion, and thought processes. Third, the brain monitors so-called pain signals coming into the spinal cord and, based on what is happening in the filtering sites, can enhance those incoming pain signals (“open the gate”), ultimately making the brain more sensitive to pain, or diminish the strength of the signal (“close the gate”). In essence, we use a simplified version of the original gate control model, retaining the general concepts, providing our own simplified illustration, and avoiding reference to complex neurophysiological terminology (e.g., interneurons, calcium channels, central sensitization, descending inhibition).

Figure 1.

Gate Control Diagram: Patient handout used to convey the concepts of the brain as a filter of so-called pain signals, emphasizing cognitive and affective mediation, and illustrating descending inhibitory (or excitatory) mechanisms [79,80].

While there is growing emphasis on the importance of patient education in chronic pain management, [37,59,66,68] patients in general practice rarely receive comprehensive biopsychosocial education as part of their treatment, and likely would not understand it unless provided in a user-friendly (jargon-free, practical) way. Considering that 49% of those in developed nations (52% in the U.S.) have basic or below basic reading skills (basic skills translate to reading at the 4^th–5^th grade level) [69] pain education, of necessity, must be simplified.

4. Simplified Pain Education and Cognitive-Behavioral Therapy

Cognitive-behavioral therapy is well established as an efficacious treatment for chronic pain [1,30,36,97], but with moderate overall effect sizes [28]. In current CBT approaches, educating the patient about pain is infused into the skills training aspects. Our clinical research team has incorporated into our treatment approach a simplified discussion of pain as a perception that is filtered by the brain, emphasizing the capacity of the brain to “open” or “close” a pain gate to process more or fewer pain signals. See Figure 1 for the illustration we use to facilitate this discussion. This is an adaptation of more complex neurophysiology pain education approaches [11,12] and it serves as our treatment rationale for psychosocial pain management, specifically, cognitive-behavioral therapy [32,79]. We emphasize the following specific points: First, contrary to what is typically understood, pain is not a sensation, but a perceptual experience in the brain that involves many filters. Second, the brain processes the sensation of pain by filtering so-called pain signals through multiple sites involved in memory, emotion, and thought processes. Third, the brain monitors so-called pain signals coming into the spinal cord and, based on what is happening in the filtering sites, can enhance those incoming pain signals (“open the gate”), ultimately making the brain more sensitive to pain, or diminish the strength of the signal (“close the gate”). In essence, we use a simplified version of the original gate control model, retaining the general concepts, providing our own simplified illustration, and avoiding reference to complex neurophysiological terminology (e.g., interneurons, calcium channels, central sensitization, descending inhibition).

In previous research [82] our laboratory evaluated two group-administered psychosocial pain treatments in patients receiving care at low-income U.S. healthcare settings. Both the cognitive-behavioral therapy (CBT) and pain education (EDU) conditions included our above described simplified gate control theory. These treatments were adapted to decrease the literacy demands of patient materials and reduce the cognitive demands of the interventions. For example, the reading level of patient written materials were lowered from 10^th grade (in original format [80]) to 5^th grade and key illustrations were used to help emphasize certain points. Further, during the treatment groups, jargon and multi-syllabic words were avoided, interactive teaching methods were used, and co-therapists provided one-on-one assistance when it appeared that group members were struggling with a concept [42]. Both CBT and EDU were provided in 10 weekly 90-minute group discussion sessions with 5–7 participants. Both treatments emphasized therapeutic alliance, group discussion, and group affiliation, and both treatments used a biopsychosocial model to provide usable information about pain, the brain, and pain self-management. CBT additionally involved active skills-building exercises (with homework), while EDU did not include skills training. Study results revealed moderate treatment effects on pain severity, and pain interference in daily functioning for both interventions, with no significant differences between treatments. EDU was better tolerated than CBT, producing fewer dropouts. Although these findings suggested that literacy–adapted CBT and EDU may both be efficacious in this population, the trial was underpowered to reliably detect differences between conditions and lacked a usual medical care (UC) comparison group, preventing evaluation of whether these therapies could augment existing medical practices.

Following additional simplifications in the CBT treatment protocol to enhance retention, (e.g., removed written homework requirements, provided CDs of session summaries to patients, additional simplifications of patient workbooks [79]), we conducted a randomized controlled comparative efficacy trial powered to detect differences between conditions (290 participants, again at low-income U.S. clinics) and added a usual medical care condition [83]. Both CBT and EDU included our simplified “gate control” education as a treatment rationale, and matching of the interventions with regard to format. The participant sample represented those with multiple disadvantages, often with the triple disparity of low-income, low primary literacy, and minority status (African American). Study completion rate (83%) showed no differences across conditions, suggesting that the additional adaptations made to CBT enhanced its acceptability. Similar to the earlier study [82], both CBT and EDU (but not the usual care condition) produced significant pre-to post-treatment decreases in mean pain intensity and pain interference, with treatment gains mostly maintained at 6-month follow-up. Comparisons of effect sizes/confidence intervals, clinically meaningful improvement rates (defined as 30% or greater improvement [26]), and minimally important change criteria [26], showed CBT having small clinical advantages over EDU on pain intensity and physical function [83]. In terms of potential harms/adverse events, 28 of all 290 participants sought emergency department treatment at some point during the trial (mostly temporary pain exacerbations). Six participants were hospitalized over night for non pain-related reasons (e.g., asthma). No events were determined to be caused by participation in the trial [83]. Both CBT and EDU showed clear advantages over and above usual medical care, answering in the affirmative that these therapies can be used to augment existing medical practices.

The unique aspect of the CBT treatment compared to EDU in the above studies is that, in addition to simplified pain neurophysiology and relevant pain self- management information, patients receiving CBT are provided with structured, guided training and hands-on practice in a variety of pain self-management skills, including relaxation techniques for stress management, activity pacing, assertiveness training, cognitive restructuring, and emotional disclosure. Each of the skills-oriented techniques taught during the CBT program are referred to as ‘gate closers’ [79] to enhance the connection for the patient between the use of cognitive-behavioral skills and the descending pain inhibitory mechanisms first proposed in the gate-control model. See Table 1 for a session-by-session breakdown of our CBT protocol as well as a comparison to the EDU protocol. Of note, this CBT program is based on an original protocol developed over decades[79,80] and supported in empirical studies [82–84].

Table 1:

Session Content in Cognitive-Behavioral Therapy (CBT) and Pain Education Protocols used in Thorn et al. (2018) [79–81].

	Cognitive-Behavioral Therapy (CBT)	Pain Education (EDU)
Session 1	Let’s Get Started: Pain is Stressful (purpose of group, and how pain and stress are related; relaxation training)	Welcome (purpose of group; chronic pain as a common problem; stress and pain connection; 4 parts of stress response)
Session 2	Manage Your Brain to Manage Your Pain (how pain works in the brain; how to reduce pain signals; relaxation)	Pain is in the Brain (how pain works in the brain; what decreases or increases pain signals in the brain)
Session 3	Getting Active (fear of pain and activity connection; training in pacing activity, scheduling pleasant activities; relaxation)	Short-term and Chronic Pain (differences between short-term and chronic pain & their treatments; pros and cons of physical activity)
Session 4	Pain and Emotions (link between pain & emotions; emotional expression training; relaxation)	Pain and Your Emotions (link between pain & emotions; chronic pain and feeling sad, mad, or scared)
Session 5	Stand Tall Talk (3 ways of communicating; assertiveness training; relaxation)	Ways of Talking to People (3 ways of communicating)
Session 6	Thoughts that Work Against You (link between thoughts, feelings, & actions; training in recognizing unhelpful thoughts; relaxation)	Talking with Health Care Workers (Importance of relationships with health care workers and maintaining good relationships with them)
Session 7	Making Your Thoughts Work for You (recognizing “red flag” words; training in changing unhelpful thoughts; relaxation)	Types and Costs of Chronic Pain (Different types of chronic pain & their treatment; multiple costs of having chronic pain)
Session 8	Master your Thoughts to Manage Your Pain (recognizing deeper beliefs; training in changing deeper beliefs, “Acting as If” exercise, coping self-statements; relaxation)	Pain Behaviors (understanding pain behaviors, what affects them, and what they might communicate; pain & behavior cycle)
Session 9	Get Better Sleep (sleep and pain connection; training in changing unhelpful sleep habits; relaxation)	Sleep (common sleep problems, normal sleep, and the sleep cycle)
Session 10	Your Pain Coping Toolbox (review of information and skills used; planning for continued pain self-management; coping with pain flare-ups)	Knowing your Pain (review of information learned; discussion of how you will use what you’ve learned going forward)

5. The Importance of Thoughts and Examples of Psychosocial Pain Intervention Techniques Targeting Them

Both CBT and EDU in the above protocols emphasize the importance of thoughts as a key process in pain-self management. An adaptation of the transactional model of stress [45] is used to provide a cognitive model of pain [80], which (simplified) is referred to as the “Think → Feel → Act” model. Traditional cognitive therapy [6] delineates categories of thought processes and targets each category therapeutically. Below is a brief description of the categories, with examples of therapeutic techniques utilized.

5a. Judgments About Stressors

Pain is stressful, as are many life challenges associated with chronic pain. An important component of the stress reaction is one’s judgment regarding the stressor. For example, the thought that a pain flare-up signals ongoing tissue damage (i.e., hurt = harm) is a judgment about the pain. Pain directs one’s attention, and is therefore compelling and interrupts function, especially if deemed to be a threat. As Eccleston (2018, pg s17) put it, for the person with chronic pain, “The lived experience …is to be in a state of defensive alert, of being urged to escape and avoid harm” [28]. One example of a clinical application using primary appraisals is to teach patients to examine their initial judgments of a situation and determine whether they are viewing it as a threat, a loss, or a challenge [79,80]. Focusing one’s attention on the judgment of the stressor (and not necessarily the stressor itself) allows them to consider working on their interpretation, often parsing out aspects that are indeed a challenge instead of a threat/loss. Another example of a clinical application of judging pain is that represented by the fear of movement/(reinjury) model and the subsequent clinical application of in vivo exposure to overcome judgments of hurt = harm [17,91,92]. Using graded exposure to physical movement, patients learn that appropriately paced physical activity using proper body mechanics does not create injury or pain exacerbations. Thus, they are able to change their judgment that physical activity causes injury to the body.

5b. Judgments about Coping Options

Many of our thoughts throughout the day arise somewhat spontaneously and provide a running commentary of environmental events. These “automatic thoughts” often occur in response to or in anticipation of pain. Examples of negative automatic thoughts might be “No one believes my pain is real” and “I won’t be able to work much longer because of my pain” [79].

Catastrophic thinking (a negative mental set about pain or potential pain) has been an imperfect, albeit robust, construct in the pain literature for decades. In hundreds of studies, catastrophizing has emerged as both a predictor of poor adjustment to pain and a specific target of intervention. An example of a catastrophizing thought taken from the well-known Pain Catastrophizing Scale [77] and included (slightly altered) in the Patient-Reported Outcomes Measurement Information System Roadmap Initiative item bank for pain catastrophizing (PROMIS) [4,13] is “I keep thinking about how much it hurts.” Classic cognitive therapy techniques such as cognitive restructuring exercises have frequently been used to (effectively) reduce catastrophizing and other negative pain-related thinking [79]. In cognitive restructuring, patients are guided to become aware of negative thoughts that work against them, and then examine whether the thought is true, partly true, or partly false. They are then trained how to come up with a more realistic, helpful, and less negative thought to “close the pain gate”. It is important to note that other techniques not targeting catastrophizing thoughts per se, such as (but not limited to) mindfulness-based therapies [20] and acceptance and commitment therapies [33], have also reported decreases in pain catastrophizing following treatment [93].

5c. Acquired “Rules” for Self and Others

Beliefs about the cause of pain and how it should be treated develop over time and are heavily influenced by cultural and social expectations. An example of such a belief is “a competent doctor who cares should find the source of the pain and fix it”. Patients with pain also develop “rules” about how they should be as a person with pain, such as “a man should work and provide for his family” [79]. Helping patients recognize the implicit “rules” they hold for themselves and others, and guiding them to assess whether these rules work for or against them can give them a sense of agency over a previously unrecognized cognitive process [79,80]. An example of a potential clinical application of an acquired belief involves the concept of perceived injustice after injury leading to chronic pain. A patient might have the belief that: “The guy responsible for this accident should be given a life sentence for taking away my life.” Therapeutic techniques that decrease anger, reduce overly harsh judgment, and increase a sense of forgiveness have all been suggested as ways of reducing perceived injustice [50,62,67,78]. Possibly, these techniques function to increase ones’ acceptance that one has a chronic condition and a willingness to engage in activities despite pain. Acceptance & Commitment Therapy, mindfulness and positive psychology techniques specifically target reduced judgment and increased forgiveness, and have also been shown to increase pain acceptance after treatment [41,65].

5d. Deepest Beliefs: Identity

In classic cognitive therapy one’s core beliefs are underlying views about the self, interacting between life events and one’s temperament [6]. Ones’ identity is often thought of as uniform and invariant, although others have suggested that we have a number of identities, often competing with each other and acting in concert with the social, cultural, and even with the internal (biological) environment [28,43]. Beck and others have suggested that one’s core identity includes both positive and negative aspects of one’s view of themselves [6]. When environmental circumstances are perceived as “positive”, more positive beliefs about the self are activated; conversely, when circumstances are perceived as “negative”, one tends to devalue oneself and hold on to negative self-constructs. Core beliefs and negative self-identity are also considered to be the most difficult to change among the types of cognitions already discussed. An example of an unhelpful core belief that may be activated with chronic pain is “I am damaged goods” [79]. Unfortunately, as pain becomes more chronic, the person with pain likely comes to develop an identity as a “disabled chronic pain patient.” This belief functions to set the individual with pain up for passivity, a “sick person” role, withdrawal, and helplessness; and the downward spiral continues.

An example of a technique that could be used to examine and possibly change negative self-identity is emotional disclosure [64,71]. Written emotional disclosure exercises have been used therapeutically with a variety of patients, including those with chronic pain [48], and focuses on teaching patients to write about (or sometimes audio record) previously undisclosed, negative, unresolved experiences. Recently, Mark Lumley and colleagues have expanded the application of written disclosure exercises to a treatment program termed emotional awareness and expression therapy (EAET) [47]. A large well-controlled trial in fibromyalgia patients comparing EAET to CBT (and an educational control) found approximately equivalent results between EAET and CBT (except that a greater percentage of EAET participants achieved >50% pain reduction), with both superior to education. Thus, EAET, in addition to more traditional cognitive therapy techniques, may hold value in helping patients consider and modify negative personal identities.

6. The movement towards identifying mechanisms and targeting those that are critical

Colleagues and I have suggested that psychosocial pain management may engage several therapeutic processes: changing what we think, i.e., cognitive content, changing how we think (i.e. cognitive processes), working with our emotions (i.e., mood states), and changing our actions to manage pain (i.e. behavior) [21,23,81]. I would argue that although traditional cognitive therapy has been previously criticized as shortsighted or superficial by targeting a change in one’s thought content (via the emblematic cognitive therapy techniques referred to as cognitive restructuring), it is possible, and even likely that techniques emphasizing cognitions also change how one thinks (i.e., alter one’s relationship to their thoughts by learning to observe one’s ever-changing thoughts without being emotionally overcome by them) [20,79]. In fact, it is quite possible that cognitive therapy techniques such as cognitive restructuring exercises do not require a change in cognitive content, but instead alert individuals to their process of thinking. As such cognitive restructuring, and other classically “cognitive” therapy exercises, may have important similarities to therapies sometimes referred to as “third wave” CBT: Acceptance and Commitment Therapy (ACT), and Mindfulness-Based Therapies.

Although different therapeutic techniques are applied, and different theorized mechanisms of change are hypothesized, there is more evidence to suggest shared, rather than treatment-specific therapeutic mediators across treatments. Consider pain catastrophizing, for example: significant reductions in pain catastrophizing have been found with biopsychosocial pain education [82,83], CBT, MBSR, and Mindfulness-Based Cognitive Therapy - MBCT [14,23,82,83], ACT [16], and even physical exercise [44,75], yoga [18,44], and Tai Chi [35]. Most of these treatments do not target changing cognitive content (e.g., MBSR, ACT), and some (e.g., physical exercise) do not target cognitions at all.

Planning treatment research that assesses mechanisms is necessary, but difficult, and results are often complex, likely reflecting reciprocal relationships between changes in hypothesized mechanism and crucial outcomes. For example in a post-hoc crossed lagged analysis of both the CBT and EDU data from our most recent RCT described above [83], we found that early changes in cognitions predicted late-treatment changes in pain intensity and interference in daily activities due to pain for both EDU and CBT. However, early changes in pain outcomes also predicted later treatment changes in cognitions [10]. Since EDU did not involve skills training or cognitive restructuring, and considering the above research reporting that a wide variety of treatments show reductions in catastrophizing following treatment, it seems likely that cognitive therapy per se is not the key procedure for cognitive change. However, cognitive change (content or process), perhaps regardless of how it is brought about, seems to be an important piece of the mechanism puzzle.

Given the few outcome differences between CBT and EDU in our study, one might conclude that EDU, which takes less cognitive effort on the part of therapist or patient and requires fewer resources on cash-strapped healthcare systems, might be a “good enough” option. In fact, Vlaeyen and colleagues made this argument after finding that pain education plus group cognitive therapy compared to pain education plus group discussion provided few differences in critical outcomes, although cognitive therapy was more expensive to administer [34,90]. It is interesting to note that these researchers also speculated that poor compliance, difficulty completing homework, and lack of individual support may have decreased the overall efficacy of the cognitive therapy program. This latter comment may speak to the importance of simplifying our psychosocial treatments.

Another approach to disentangling treatment mechanisms is to move beyond examining average outcomes between or across treatment groups and to examine moderators of treatment efficacy. That is, do differences among patients predict whether they will get more or less benefit out of a particular treatment? (i.e., heterogeneity of treatment effects). In another post-hoc analysis of the comparative efficacy trial conducted by our group [83], we looked at common demographic and potential cognitive moderators of treatment outcome differences across conditions. As expected, few meaningful moderation effects were found in the usual medical care condition. Interestingly, neither sex nor minority status moderated differences between any of the conditions. However, those patients with lower education, literacy, and working memory benefitted more from CBT than EDU [27]. We believe this unexpected effect can be attributed to our team’s multi-year effort to simplify our CBT patient materials and therapeutic process, removing the requirement for written homework, and providing enhanced individual support within the structure of the group for those with lower literacy levels. Thus, it seems that those with higher literacy and cognitive skills were able to take the biopsychosocial pain education information and apply it to their daily lives, whereas those with lower literacy and cognitive function benefited more from the structure and support provided by specific skills training without the burden of written homework. This interpretation has some support from the qualitative analysis of post-treatment interviews in our earlier trial [22], in which some EDU participants said they were able to “take the information and run with it”, while some CBT participants commented that specific training and practice (e.g., behavioral pacing) was useful in helping them engage in pain self-management behaviors.

7. Future Directions and Conclusion

The research findings reported above hold potential promise for implementing biopsychosocial treatments in many low-income settings and developing nations with patients previously thought to be inappropriate for a cognitively-oriented approach. CBT is thought to require abstract reasoning, comfort with reading and writing, and written homework compliance. It is important to underscore that our results suggest that those with lower reading and cognitive function did better with the structure of CBT (as simplified), while higher functioning individuals benefitted from either simplified EDU or CBT. Our findings suggest that simplifying our psychosocial techniques would reach a wider swath of individuals than the typically more privileged highly educated or wealthy nations. Furthermore, the structure and guidance provided by simplified CBT skills training seems to be a better option than EDU when limited educational attainment and literacy are concerns. An important caveat is that both CBT and EDU as delivered in our trial relied heavily on motivational approaches via strong therapeutic alliance and group coherence to provide an accepting and collaborative learning environment. This emphasis, including a treatment rationale that is understandable and makes sense, likely instilled positive outcome expectancy. Furthermore, both treatments involved multiple sessions (10) over the same number of weeks with adequate time provided for interactive learning and discussion (90 minutes). Patients received culturally tailored and literacy-adapted workbooks and audio CDs to facilitate comprehension. Efforts are underway to adapt and administer the CBT protocol to disadvantaged populations in Australia [89], and similar efforts to adapt and simplify related treatments (e.g., pain neuropsychology education) are being tested in Nepal [73].

In conclusion, I would argue that all pain is brain mediated and therefore, all pain is organic. I think it is unlikely that a biomarker for objective pain will be found despite the field’s tenacity to identify one [19], since such a search suggests that some pain is real and some is not. Assessing and treating only one aspect of pain (i.e., tissue damage) has likely led to iatrogenic harm and much suffering. Further, our patients (and society in general) have been trained to expect that pain can be eradicated by medical professionals. Patients have had little access to pain self-management training, nor have they been given relevant, useful (and user-friendly) information about the brain – the seat of this perception we call pain. This review highlighted some biopsychosocial interventions that conceptualize pain using a cognitive model, offering unique research assessing the impact of simplifying our message and our psychosocial treatments. Continuing interdisciplinary collaboration in research and treatment will always be in the best interest of our patients.