Access to the PPM Journal and newsletters is FREE for clinicians.
9 Articles in Volume 8, Issue #5
Chronic Pain and Substance Abuse
Eye Screening and Intractable Pain Management
Pain and the Brain
Postherpetic Neuralgia Pain and Laser Acupuncture
Prolotherapy for Golfing Injuries and Pain
Proposed Models of Fibromyalgia Sub-types
Realistic Pacing of Pain Patients’ Activities
Safe Analgesic Use in Patients With Renal Dysfunction
Superior Pharyngeal Constrictor Muscle Dysfunction

Pain and the Brain

A synthesis and summary of the current understanding of brain response to pain learned from functional magnetic resonance imaging (fMRI).

Pain is a complex biopsychosocial process that impacts millions of Americans each year and results in human suffering and significant costs to society via health care expenditures and lost productivity.1,2 With the advent of functional magnetic resonance imaging (fMRI), our understanding of the underlying physiological/anatomical mechanisms of pain, as well as the impact of proven interventions for the treatment of pain, can be more fully understood. This new technology provides the ability to identify a variety of cortical and subcortical structures involved in the cerebral response to nociceptive stimuli in humans. This article aims to synthesize and summarize the current literature on what has been learned about pain through fMRI.

Basic Description of fMRI

There are several different brain imaging techniques that can be used for the study of pain: positron emission tomography (PET),3 single photon emission computerized tomography (SPECT),4,5 Magnetoencephalography (MEG),6 functional magnetic resonance imaging (fMRI),7 and others.8 The advantage of using brain imaging tools is that they are non-invasive, and can be applied to both normal and abnormal subjects to study dynamic brain activities. Among these tools, fMRI is currently the most widely used.

The fMRI was developed on the basis of MRI—a discovery by Felix Bloch and Edward Mills Purcell, who were awarded the Nobel Prize in Physics in 1952 for this discovery. The principle of MRI is that atomic nuclei will rotate with a frequency that is dependent on the strength of the magnetic field. In the absence of a strong magnetic field, hydrogen nuclei (protons) are randomly aligned. When the strong magnetic field is applied, the hydrogen nuclei wobble or precess about the direction of the field. Therefore, when a human body is placed in a large magnetic field, many of the free hydrogen nuclei align themselves with the direction of the magnetic field. A radio-frequency (RF) pulse is then delivered whose energy can be increased if they absorb radio waves with the same frequency (resonance). When the RF pulse stops, the nuclei return to equilibrium or relaxed state and the nuclei realign themselves at their net magnetic moment. When the atomic nuclei return to their previous energy level, radio waves are emitted to be measured, processed, and reconstructed to obtain 3D grey-scale MR images. Paul C. Lauterbur and Peter Mansfield pioneered the application of MRI in medical imaging and led to their Nobel Prize in Physiology and Medicine in 2003.

To enhance the signal, intravenous injection of gadolinium was first used in the fMRI as a contrast agent in the visual cortex and evoked by flashing light.7 However, it was soon realized that the body has its own natural contrast agent—deoxygenated hemoglobin.9-11 Most fMRIs now rely on the blood oxygenation level dependent (BOLD) effect, which is based on the increased neuronal firing in response to a stimulus that will induce hemodynamic changes and ultimately modify the magnetic field to increase the MRI signals.12 It is thought that an increased metabolic demand, due to increased neuronal activity, results in an increase in blood flow beyond metabolic needs, such that the final ratio of deoxyHb/oxyHb actually is reduced. It is the reduction in deoxyHb that alters the magnetic field properties and produces the increased MRI signal.13 On the other hand, by using visual system circuitry to elicit synaptic activity, a strong correlation was observed between local field potentials and change in brain oxygen concentration without generation of action potentials.14 The advantages of fMRI include its non-invasiveness, good spatial (down to 1 to 2 mm) and temporal (100s of ms is possible) resolution. However, fMRI has its limitations such as expense, availability due to time-sharing on a clinical scanner, restriction to only those having no metallic devices in their bodies, and loud noise.

As will be reviewed in greater detail later in this article, fMRI has been conducted to investigate brain mechanisms underlying both chronic and acute pain. The fMRI has been used to study stimulus-related responses, such as noxious electrical stimulation of the skin or peripheral nerve,15-17 noxious heat or cold,18-23 mechanical,24 chemical,25,26 or visceral27 stimuli. The fMRI has also been used to image allodynia in complex regional pain syndrome,28 and patients suffering from neuropathic pain.29 Several brain areas can be activated. They include the SI, SII, anterior insula, ACC, thalamus, and cerebellum.

The relationship and interaction of pain, attention, and anticipation has been demonstrated to activate slightly different areas of the brain.15,22 For example, anticipation of pain activated the anterior ACC, whereas the pain itself activated the posterior ACC. Further dissection on this line demonstrated that the posterior insula/secondary somatosensory cortex, the sensorimotor cortex (SI/MI), and the caudal ACC were specific to receiving pain, whereas the anterior insula and rostral ACC activation correlated with individual empathy scores when the subjects watched their loved ones receiving pain stimuli.30 It was suggested that the modulatory effect of expectation on pain transmission might involve activation of descending modulatory systems.31 On the other hand, with the combination of psychophysical assessment and fMRI, pain-related activations have been obtained in parallel psychophysical sessions,18 or during the imaging sessions,15,23,26 in order to separate those activations due to the mere presence of a stimulus (due to attention) from those related to the subjects’ actual sensory experiences.

Despite the technical advantages of the fMRI, the notion of whether fMRI is a modern phrenology is under debate.32-34 One should be cautious of its potential abuse,35,36 especially when it is used in dissecting the cognitive and emotional mechanisms, because cognitive function is a “moving target.” “Ask a person a question once, and it is a different person to whom you repeat that question the next minute or the next day.” A good fMRI study is difficult to design, conduct, analyze, and interpret; special expertise is needed to design a methodologically-sound study.

Brief Overview of the Physiology of Pain

Prior to a discussion of more specific fMRI findings, a review of what was known about the physiology of pain prior to the advent of fMRI is warranted. At the level of the peripheral nervous system, myelinated Aδ fibers play a significant role in the transmission of sharp pain, while slower, unmyelinated C-fibers transmit dull or burning pain sensations. Pain signals travel along a combination of these fibers and enter the central nervous system at the dorsal horn of the spinal column.37 Within the dorsal horn of the spinal column, pain signals are modified before ascending to different brain regions.38

With the introduction of the Gate Control Theory of Pain by Melzack and Wall,38 a consensus grew that pain is a complex psychological-perceptual process. Subsequently, the experience of pain was hypothesized by Melzack and Casey,39 to consist of three dimensions: a) sensory-discriminative; b) affective-motivational; and c) cognitive-evaluative. Consistent with the hypothesis of different dimensions of the pain experience, fMRI studies have sharpened our understanding about the different pathways and neural networks thought to be responsible for the different dimensions of pain.40 Moreover, ongoing research regarding the processing of pain in the dorsal horn of the spinal column has increased understanding of perceptual experiences— such as allodynia and hyperalgesia—and will be discussed later in the article.41

After pain signals have entered the central nervous system via the dorsal horn of the spinal column, pain signals travel via several tracts to the brain. Spinothalamic tract neurons project onto the thalamus, which projects onto the primary and secondary somatosensory cortices. The primary somatosensory cortex is located in the parietal lobe in the postcentral gyrus, and it is the main receptive area for touch and pain. The primary somatosensory cortex receives the majority of the projections from the thalamus (which is the primary “relay” station of afferent sensory signals). A sensory homunculus, which represents different parts of the body, maps onto the postcentral gyrus and corresponds to similar areas of the precentral gyrus—the location of the motor cortex. The primary somatosensory cortex contains cells that project to the secondary somatosensory cortex. Although the presence of the secondary somatosensory system has now been fully established, the precise function of the secondary somatosensory cortex remains somewhat unclear. However, evidence from fMRI studies suggests that the primary and secondary somatosensory cortices are crucial in the sensory-discriminative dimension of the pain experience, such as processing the intensity and location of pain.

The affective-motivational dimension of pain appears to correspond with the pathways in which ascending nerve fibers project onto the thalamus, hypothalamus, and limbic system. The limbic system, with its projections to the prefrontal cortex, plays an important role in emotion, memory, and attention. The anterior cingulate cortex and the insula cortex, also known as the insular cortex, repeatedly have been shown to be of importance in the pain experience. The anterior cingulate cortex consists of nerve cells anterior to the corpus callosum and is involved in several areas of functioning, including the regulation of autonomic nervous system, reward anticipation, decision making, and emotion. Bush and colleagues42 divided the function of the anterior cingulate cortex based on location:

  • the anterior portion of the anterior cingulate cortex is hypothesized to be important in executive functioning
  • the posterior important in evaluative processes
  • the dorsal important in cognitive functioning
  • the ventral component important in emotional functioning

Furthermore, the anterior cingulate cortex is connected to several brain areas, including the prefrontal cortex, parietal cortex, and the motor system. The anterior cingulate cortex is hypothesized to be important when effort is needed to perform a task.

The insula cortex is a critical component of the limbic system, which functions in the regulation of emotion and memory. Damasio43 hypothesizes that the insula cortex associates visceral states with emotional experience to create an associate-experience among bodily sensations and emotions. The rostral insula cortex is related to autonomic and limbic functioning, while the caudal insula cortex is related more to motoric functioning. The insula is thought to play a role in both the sensory-discriminative, as well as the affective-motivational, dimensions of pain.41

In addition to the areas mentioned above, projections from the dorsal horn of the spinal column ascend to, and descend from, other areas of the brain, including the medulla, reticular formation, and periaquaductal grey matter (which are hypothesized to also impact the perception of pain). For instance, Mayer and Price44 performed a study in which stimulation of the periaquaductal grey matter decreased pain. As use of fMRI continues, there will no doubt be an increase in our understanding of the modulatory role of these descending systems.

Overall, fMRI findings have revealed the importance of the primary and secondary somatosensory cortex in the processing of the sensory-discriminative dimension of pain (the lateral pathway), and the anterior cingulate cortex and insula cortex in the processing of the affective-motivational dimension of pain (the medial pathway); see Figures 1 and 2. In general, the more intense the pain, the greater the activation of these four regions. Furthermore, fMRI studies provide support for the hypothesis that pain is processed within a pain matrix or brain network consisting of the areas described above, rather than in discrete, independent areas of the brain.41

Figure 1. Pain matrix; lateral and medial pathways are illustrated (after Treede RD et al45). The lateral pathway is thought to correspond to the sensory-discriminative component of pain while the medial system is thought to correspond with the affective-motivational component. Pain signals travel from the periphery, enter the dorsal column of the spinal cord, and are sent to the thalamus and on to the somatosensory cortices (lateral pathway) or insular and anterior cingulated cortex (medial pathway).

Figure 2. fMRI images of areas associated with the perception of pain46 (used with permission of Karen D. Davis, PhD). A consistent finding among neuroimaging studies is increased activation of the primary and secondary somatosensory cortices during pain, which is related to the sensory-discriminative aspect of the pain experience. The anterior cingulated cortex and insula cortex are thought to correspond to the affective-motivational aspect of pain.

fMRI and Pain

The inherent subjective reality of pain creates challenges to the many professionals who devote their careers to bringing relief to those who chronically suffer from it.47 However, new findings from fMRI studies help us to better understand the effect of interventions on the pain experience, and they may allow us to refine tactics to better manage pain. For example, the impact of distracting techniques during painful stimuli has been investigated using fMRI, and results generally support greater activation of the medial system and decreased activation of the lateral system during the implementation of distraction techniques.41 According to Valet and colleagues,48 evidence suggests that pain modulation from descending fibers is engaged by increased activity in the prefrontal cortex and anterior cingulate cortex, elements crucial in the affective-motivational dimension of pain. The affective-motivational correlates of pain have also been investigated more precisely with fMRI. For instance, Ploghaus and colleagues22 examined the anticipatory aspects of the pain experience using fMRI, and they found greater activation in the rostral anterior insular cortex and medial prefrontal cortex when a subject was anticipating pain, and greater activation in the caudal insula and anterior cingulate cortex when a person experienced pain.

Furthermore, fMRI studies allow us to more fully investigate phenomena that have previously been relatively poorly understood. For instance, Wager and colleagues49 examined the placebo effect using fMRI and concluded that the rostral anterior cingulate cortex and brainstem showed increased activation in both the placebo and analgesic condition. In a fascinating study of empathy, Singer and colleagues30 examined individuals who personally experienced a physically painful stimulus, and compared their brain activity to observing a loved one undergoing a physically painful stimulus. Similar patterns of brain activity in the affective-motivational system were observed in both conditions: when a person underwent a painful stimulus or when a person observed a loved one undergoing a painful stimulus. Namely, in both conditions, an increase in activity in the anterior insula and rostral anterior cingulate cortex, as well as in areas of the brainstem and cerebellum, were found. Furthermore, the amount of activation of the anterior insula and anterior cingulate cortex correlated with scores on measures of empathy. Mackey and coworkers50 have also shown a relationship between empathy pain and fMRI indices. Interestingly, Eisenberger and colleagues51 found greater activation of the anterior cingulate cortex when someone experienced emotional pain—specifically, emotional distress over being excluded. The prefrontal cortex was also found to play a role in the processing of emotional pain.

A Tool for Better Understanding Different Pain Syndromes

The use of fMRI technology will be of great importance in learning more about differences in biopsychosocial underpinnings of different pain syndromes. This, in turn, will lead to better “tailored” treatment for patients with different pain syndromes. For example, just on a basic level, Apkarian and colleagues52 have found differences in the brain-activity response to pain in healthy subjects with no history of chronic pain relative to chronic pain patients. Also, as mentioned earlier, fMRI has been used to image allodynia in complex regional brain syndrome,28 as well as neuropathic pain.29 In addition, Gracely and colleagues53 used fMRI to demonstrate augmented pain processing in fibromyalgia patients. This was also found in low back pain patients.54 Mackey and colleagues have also found similar findings in temporomandibular joint pain patients (Mackey, personal communication). Finally, Cutrer55 has reviewed the use of functional imaging in primary headache disorders.

Thus, there is a growing amount of work demonstrating areas of the brain specifically associated with certain pain syndromes. The significance of this work is not only limited to better understanding potentially important cortical areas involved in the etiology of different pain syndromes, but also an objective tool to use in the evaluation of treatment outcomes. The important goal will be to evaluate the validity of its use in documenting actual clinical improvements as a result of certain known effective treatment methods. This is not to say that fMRI will necessarily be the absolute “gold standard” measure, but simply one to be used with other biopsychosocial measures to most comprehensively evaluate pain.

fMRI and Interventions

The behavioral medicine techniques of distraction, relaxation training, hypnosis, and biofeedback have been utilized for over 40 years to help patients manage their pain. Although the effectiveness of behavioral medicine techniques to manage pain has been repeatedly demonstrated, neuroimaging studies and, more recently, fMRI studies allow for a more detailed understanding of how these interventions impact the processing of pain.56 The section below describes what has been learned using fMRI, as well as other neuroimaging, in patients with pain during various behavioral medicine interventions.


The Executive Committee of the American Psychological Association – Division of Psychological Hypnosis57 defines hypnosis as “a procedure during which a health professional or researcher suggests that a patient or subject experience changes in sensations, perceptions, thoughts, or behavior…” After the induction phase, which typically consists of a relaxation or fixation, there is a suggestion phase in which the individual who is under a hypnotic state is open to the suggestions of the hypnotherapist. After continuously being given cues to remain in the deep state during the suggestion phase, the individual in the hypnotic state is “taken back up” and guided out of the deepened state. This type of intervention has been utilized with a wide variety of medical and psychological conditions ranging from depression and stress to burn injuries and gastro-intestinal disorders. There have been many studies assessing the relative efficacy of hypnosis within chronic pain patients.58

Faymonville, Boly, and Laureys59 have utilized positron emission tomography (PET) to anatomically differentiate hypnotic suggestion from mere distraction and relaxed states. A great deal of research consistently supports the notion that the anterior cingulate cortex plays a prominent role in the modulatory effect that hypnosis exerts in different pain conditions.59,60 Other studies utilizing fMRI have investigated the relationship between pain and hypnosis. For example, many studies have utilized fMRI to address the structures in the brain that the sensation of pain is located in or around. In a study that compared noxious heat and hypnotically suggested pain within the same sample, the researchers found that the hypnotically induced pain activated similar areas compared to the administered painful stimulus; however, the magnitude of activation in the hypnotically induced pain was somewhat attenuated.61

In addition to localizing certain pain processing areas, many studies address the potential analgesic effects of hypnosis as a treatment modality.62 For example, researchers found that clinical hypnotic analgesia, accompanied with thermal pain, activated areas of the anterior basal ganglia and left anterior cingulate cortex at a much higher rate than thermal pain without the accompaniment of hypnosis.63 Furthermore, in addition to these activations, a reduction in pain was reported by individuals undergoing hypnosis training. These sites are qualitatively different than the structures involved in counterstimulation and distraction. The researchers suggest that the activation of these unique areas provide clues as to the inhibitory pathways involved in hypnotic analgesia.

Finally, it should be noted that, although the findings suggest that there can be direct physiological consequences by utilizing various hypnosis procedures, there are still certain individuals who are more or less susceptible to enter a hypnotic state. Using the Stanford Hypnotic Susceptibility Scale Form C,64 many of the research studies use quantified cut-off points and include individuals who are able to enter a hypnotic state relatively easily. This notion begs the question of whether hypnotic suggestion is a viable treatment modality for any type of chronic pain patient. This issue was addressed by researchers in which they attempted to improve hypnotic susceptibility utilizing EEG neurofeedack.65 After individuals were trained in EEG neurofeedback, they were able to raise hypnotic susceptibility on an often-used measure by training individuals to change the theta/alpha wave ratio that their central nervous system exhibits.

Biofeedback and Stress Management

Numerous studies have shown that stress can have a mediating effect on pain conditions through the relationship between the limbic connections and the pain processing neuromatrix. For instance, Stoeter et al66 showed that when chronic pain patients are subjected to a painful stimulus (e.g., pin-prick), as well as cognitive stress (e.g., sequencing symbols under time pressure), the superior temporal gyrus has similar activations in both cases. Furthermore, in another study, researchers found that relaxation training utilizing biofeedback and the subsequent decrease in the electrodermal response was found to activate the anterior cingulate cortex, insula, somatosensory cortices, and the amygdala, all structures within the brain that are functionally related to the experience of pain.67

Currently, fMRI is beginning to be used as a biofeedback modality. The function of the anterior cingulate cortex varies across many areas of the brain. However, as mentioned, the anterior cingulate cortex is generally thought to play a mediating role among cognition, perception, and motor control with regard to emotional states. DeCharms et al68 conducted a highly innovative study utilizing real-time fMRI (rtfMRI) to examine the rostral anterior cingulate cortex (see Figure 3). As the name implies, rtfMRI measures and provides feedback of brain processes as they occur. DeCharms and colleagues’ earlier research69 demonstrated that individuals have the ability to gain voluntary control over activation over a specific area of the brain. Extending their earlier work, DeCharms et al68 trained healthy volunteers to increase or decrease activation of the rostral anterior cingulate cortex. Perception of pain during an acute thermal stimulus changed in the expected direction (e.g., decreased pain sensation corresponding to decreased activation of the rostral anterior cingulated cortex). Furthermore, patients with chronic pain were able to decrease levels of ongoing pain by decreasing activation of the same region of the brain. This study has tremendous implications for the ways in which proven behavioral medicine techniques impact the perception of pain.

Figure 3. Changes in activation of the anterior cingulated cortex in individuals undergoing real-time fMRI training. Row (A) demonstrates activation comparing the first and last training session. Row (B) demonstrates repeat analysis comparing the post-test session to the first training session. Perception of pain during an acute thermal stimulus changed in the expected direction (deCharms et al68; used with permission).

In addition to the groundbreaking work of deCharms and colleagues,68,69 a number of other investigators have found similar success in the use of rtfMRI in individuals’ ability to develop self-control of different areas of the cortex.70,71 It should be kept in mind, though, that rtfMRI is still in its infancy and, as Andrasik and Rime72 have noted, further progress is still greatly needed in areas such as: the standardization of methodologies across investigators/laboratories; controlling artifacts such as head movements and respiratory patterns; and better understanding of individual differences in brain structure and processes. Henning and colleagues73 had earlier provided a comprehensive review of these methodological issues, as well as potential clinical applications. Nevertheless, progress is being made, such as developing optimal rtfMRI training methods based on operant learning principles (e.g. Bray et al74).

The Placebo Effect

As Gatchel75 has highlighted, clinicians should always be aware of a potentially powerful adjuvant process for pain relief—the placebo effect. He has reviewed research that demonstrates the role of the placebo effect in a number of different areas, including its significant role in the relief of symptoms by chemically inert drugs; as an active ingredient in psychotherapy and behavioral therapy, especially when anxiety is being treated; and in the reduction of pain. The question of whether the placebo effect is a “real” phenomenon is now being answered in the affirmative on the basis of fMRI research. For example, as discussed earlier, Wager and colleagues49 have examined the placebo effect using fMRI and concluded that the rostral anterior cingulated cortex and brainstem showed increased activation in both the placebo and analgesic condition. Other investigators have also isolated fMRI indices of the placebo effect.76,77


As reviewed by Gatchel, Peng, et al,78 the biopsychosocial approach to chronic pain has spurred significant new advances in assessment and treatment, as well as better understanding etiological factors. One important new contribution has been the development of technology—such as brain imaging—that provide new insights into brain-pain mechanisms. In the present article, we have provided a broad overview of this exciting new field of fMRI and pain. Besides its use for the assessment of pain-related syndromes, it also adds a new intervention to our pain-management armamentarium. It is another potentially important behavioral medicine intervention which can be utilized alongside other therapeutic methods in a comprehensive, interdisciplinary manner. Although it has been known that behavioral medicine strategies are effective tools for pain management,1 fMRI studies also allow us to gain better understanding as to why these techniques are effective. As practitioners work toward advancement with regard to the care of chronic pain patients, it will be important to utilize treatment interventions that give patients an increased sense of control. The increased use of fMRI and other neuroimaging therapies is certainly a “stepping stone” toward future advancements.

Last updated on: December 20, 2011
close X