What if psychological treatments for pain are more powerful than they’re given credit for?
Tor Wager, PhD, a cognitive psychologist and head of Dartmouth College’s Cognitive and Affective Neuroscience Lab (CANLab), thinks they may be. Wager and his colleagues have found that people’s experiences of pain are heavily influenced by their thoughts, beliefs and sociocultural contexts—a discovery that might not surprise your average psychologist. But what makes CANLab’s work different is the researchers’ use of cutting-edge neuroimaging and machine learning methods to ground these discoveries in the circuitry of the brain.
Take, for example, a recent study by Elizabeth Reynolds Losin, PhD, now an assistant professor of psychology at the University of Miami. As a CANLab postdoc, she became interested in the contradictions surrounding racial and ethnic differences in pain. In research and clinical settings, African Americans report more sensitivity to pain than other racial groups. And yet, African Americans’ pain is notoriously undertreated, and research has found that medical professionals often believe that African Americans feel less pain than white patients.
Wager, Losin and their colleagues recruited 30 African Americans, 30 white non-Hispanic Americans and 30 Hispanic Americans to have brain scans with fMRI while being subjected to painful heat. As expected, the African American participants reported more pain for a given level of heat than the white or Hispanic participants.
The fMRI data allowed Losin and Wager to find out why. The African American participants’ greater sensitivity to pain was correlated with increased perceptions of discrimination, as well as decreased trust in the experimenter (a white man). These feelings of discrimination and mistrust, in turn, were linked to higher activity in frontostriatal regions of the brain, activity that was also indicative of greater subjective ratings of pain (Nature Human Behaviour, online first publication, 2020opens in new window).
“Pain is more threatening when it’s given in the context where you don’t feel as safe or if controlled by a person you don’t trust,” Wager says.
Understanding how complex psychological phenomena like trust can affect a basic process like pain is the linchpin of Wager’s work in the CANLab, which relocated to Dartmouth from the University of Colorado Boulder (CU Boulder) in 2019.
Wager’s goal is to better understand and provide new treatment options for chronic pain, which affects approximately 50 million American adults. For many, the real agony they feel is not associated with physical damage in the body, Wager says. To understand the actual sources of pain, he and his team are studying basic emotional processing, responses to aversive sounds and sights, sociocultural influences and the placebo effect. They’re also studying pharmacological and psychological treatments.
“If we understand these things on a physiological level, if we could say, ‘Absolutely there is nothing wrong with your back, but here’s this circuit in your brain that is causing learned-fear avoidance that is amplifying your pain signals,’ we could know how to really help individual patients get better,” Wager says.
Placebo power
This focus on pain has personal resonance for Wager. He was raised in the Christian Science tradition, a religion that holds that thoughts and belief are the source of all healing. Though he doesn’t hew to those beliefs himself, he says he always found the idea of the power of belief to be fascinating. And pain is an ideal lens through which to study thoughts and beliefs, he says, because scientists know a great deal about the physiology of pain in the peripheral nervous system, such as the mechanics of pain receptors and basic nociception, but not as much about what is happening in the brain. “The neuroscience of pain above the neck is really a new field,” he says.
Wager’s first major contribution to that field was a seminal study on the placebo effect, which he completed as a graduate student at the University of Michigan. He and his colleagues applied small electrical shocks and heat to participants’ skin while using fMRI to track brain activity. The researchers applied a cream to all the participants’ skin prior to the painful stimuli, telling some of them that it was an inert lotion and others that it was a highly effective painkiller. The cream was a placebo, but patients who believed it to be a real painkiller reported less pain than control participants.
That pain relief was reflected in their brain activity: During the application of the painful stimulus, placebo-condition participants showed reductions in brain activity in painrelated brain structures such as the anterior cingulate cortex, the contralateral insula and the contralateral thalamus. The placebo also increased activity in the dorsolateral prefrontal cortex and orbitofrontal cortex during the anticipation phase of the experiment, when participants were steeling themselves for pain (Science, Vol. 303, No. 5661, 2004opens in new window). Follow-up research found that placebo treatment increased the release of endogenous opioids during pain in the orbitofrontal cortex, the right amygdala and the right anterior cingulate cortex, all pain-related regions (Proceedings of the National Academy of Sciences, Vol. 104, No. 26, 2007opens in new window). The researchers had shown how belief alone could alter fundamental organic processes in the brain.
“It’s really a pure manipulation of thought, belief and context,” Wager says of placebo treatments. “You manipulate psychology and then you get an effect on body outcomes and brain outcomes.”
Pinpointing pain in the brain
But to better understand these processes, Wager and his team have had to develop an objective way to detect pain in the brain. As the placebo studies elegantly illustrate, pain is highly subjective. One person might give a stimulus a rating of 10 on the pain scale while another person would rate it as 5. And, more to the point, what someone rates as 10 on the pain scale on one day might feel like 5 to them on a better day. More rested, less stressed, more comfortable with the people you’re with? You’ll probably rate any given noxious stimulus as less painful.
To dig into the specific reasons that pain feels the way it feels to a particular person at a particular moment, Wager and his team are seeking to identify brain biomarkers for pain by using machine learning to find patterns of brain activation in fMRI data that signal pain. Machine learning can pick out patterns far too complex for human scientists alone to pluck from a large set of data, says Philip Kragel, PhD, a postdoctoral researcher in Wager’s lab who is based at CU Boulder. The goal is to use these pain biomarkers in research to understand how pain processing works and to test interventions for pain.
In 2013, Wager and his colleagues published their first paper describing a successful pain biomarker, which they call the neurologic pain signature (The New England Journal of Medicine, Vol. 368, No. 15, 2013opens in new window). This signal consists of a complicated pattern of activation across many brain regions associated with pain and sensation, including the secondary somatosensory cortex, the insula and the anterior cingulate cortex. The signal could discriminate between painful heat and nonpainful warmth with 93% sensitivity and specificity and between physical and social pain with 85% sensitivity and 73% specificity. Administration of an opioid painkiller caused the neurologic pain signal to subside.
Much of the lab’s work now uses this neurologic pain signal as a benchmark in research studies to understand its participants’ painful experiences. The researchers are also continuing to use machine learning to try to expand their suite of biomarkers. Kragel’s work in Boulder focuses on defining patterns of activation that mark particular emotional experiences. Marta Čeko, PhD, a postdoctoral researcher also based at CU Boulder, is on the hunt for brain activity that is common to aversive experiences more generally, whether pain, unpleasant sounds or unpleasant visual stimuli. The findings may translate to a deeper understanding of emotional experiences as well.
“Anxiety, depression, pain, all of those have some kind of dysregulation in affective processing, which includes aversive processing,” Čeko says. “If there are circuits that are common to those three, we can find treatments that target the common denominator.”
The psychological side of pain
The neurologic pain signature has opened up a cascade of new investigations into the intersections of pain, emotions, expectations and experience. Wager and his co-authors have found, for example, that when people expect a lot of pain, their neural pain responses to a given painful stimulus are heightened, leading to a greater sensation of pain and even higher expectations of future pain—an ominous positive feedback loop (Nature Human Behaviour, Vol. 2, No. 11, 2018opens in new window).
Many lab members are particularly motivated to tackle chronic pain, which is defined as pain that lasts for more than three months and which may or may not have a detectable physical cause. Anxiety and depression are risk factors for chronic pain, hinting that aversive learning circuits in the brain contribute to the problem. Sociocultural factors may matter, too. In the study of pain differences by race and ethnicity led by Losin, the patterns of brain activity seen in the pain-sensitive African American participants were similar to the patterns seen in learned pain avoidance and in the brains of patients with chronic pain (Nature Neuroscience, Vol. 17, No. 11, 2014). Adverse life experiences and chronic stress are known to be linked to chronic pain, Losin says.
“Perhaps here, discrimination as a chronic stressor is causing these brain regions to become responsive to pain in a similar way as they are to chronic pain patients,” she says.
So far, the neurologic pain signature is used alongside a painful stimulus to quantify the intensity of the pain a person perceives from that physical sensation. Wager and his team are now working to identify patterns of responsiveness that can quantify chronic pain that may not have an obvious physical trigger. “It’s going to be challenging,” he says.
Meanwhile, he and his colleagues are already working on testing interventions for chronic pain. In a study being prepared for peer review, participants who had been in pain for an average of 10 years participated in a randomized clinical trial with two treatments: an open-label placebo injection into the back, with participants’ full knowledge that they were receiving a placebo, or four weeks of twice-weekly sessions of a novel psychological treatment developed and delivered by Wager’s collaborator, Los Angeles psychotherapist Alan Gordon, LCSW, focused on reducing fear and anxiety around pain. The goal is to test the novel psychological treatment for pain as well as whether placebos that patients know are placebos reduce pain over long time periods.
So far, participants in the placebo group have shown a significant reduction compared with a waitlist group, Wager says. This improvement was modest but was maintained at a three-month follow-up. Wager has been documenting the participants’ subjective responses to the treatment as well as their neurological changes, and results from the psychotherapy treatment are forthcoming.
One possible limitation of many current psychological approaches to pain is that patients are told that the treatments won’t change the pain, only help them cope with it. That may be untrue and ultimately counterproductive, given the results seen in the CANLab so far, Wager says. “If you’re delivering a great psychological treatment for pain, based on acceptance but with a negative expectation, you’re cutting it off at the knees,” he says.
But to truthfully offer patients hope that psychological treatment won’t just help them cope with pain but actually erase the agony, Wager and his team will need to be able to identify the patients who can benefit. That means understanding pain circuitry well enough to rule out people with serious bodily problems that are causing their pain.
“The real action here that is important for people is right at the intersection of this interplay between physiology and psychology,” Wager says. “That’s what we’re studying.”
“Lab Work” illuminates the work of psychologists in research labs. To read previous installments, go to Monitor Digital and search for “Lab Work.”
