A historical perspective on our understanding of fibromyalgia

In 2017, the International Association for the Study of Pain formally acknowledged that there was a third descriptor/mechanism of pain, termed nociplastic pain. This move went largely unnoticed by non-pain researchers. But for those in the field of pain that had been studying these conditions, or individuals with these conditions, this was a watershed moment. This represented formal acknowledgement that the science had evolved to show that in pain conditions such as fibromyalgia (FM), the pain and other symptoms that these individual experiences are very real and are likely coming from the central nervous system (CNS) rather than from ongoing peripheral nociceptive input.

This notion that all pain had to have a peripheral cause was firmly held in both the field of pain and more broadly in clinical medicine. In the pain field, nearly all preclinical pain researchers were of the belief (many still are) that in order to experience pain, there must be activation of a nociceptor in the periphery. Most animal models of pain intentionally ignore studying cortical or subcortical influences on pain. Not surprisingly then, massive amounts of preclinical animal pain research have taught us a tremendous amount about the mechanisms involved in nociception (ie, the processes that involve activation of a peripheral nociceptor) but much less about pain, and very few drug targets identified by these peripherally focused animal models have been successfully translated into new analgesics.

In clinical medicine, there has been a parallel focus on peripheral drivers of pain. In subspecialty training, there is typically an intense focus on studying one region of the body. When individuals have pain in that region of the body or tissue, subspecialists carefully interrogate this region with both the history and physical exam and then have an increasingly complex set of imaging studies, biopsies, etc that can be used to determine if there is peripheral damage or inflammation in that region of the body. After these thorough clinical investigations—in nearly every clinical subspecialty—providers came up with one or more labels for individuals who had pain—but no demonstrable inflammation or damage that should be causing that pain. In rheumatology, our such syndrome was fibrositis. But gastroenterology had spastic colitis, urology had interstitial cystitis, dentistry had TMJ, gynaecology had endometriosis and so on.

Nearly, all of these conditions were formally renamed by subspecialty experts beginning in the 1970s, with the new names purposely connoting the fact that evidence suggested strongly that this condition was not characterised by peripheral damage or inflammation driving the pain. Often the suffix is replaced because of an absence of inflammation. Fibrositis became FM, spastic colitis became irritable bowel syndrome (IBS) and interstitial cystitis became bladder pain syndrome. These foundational shifts in our understanding of each of these conditions took decades but one by one came to the same conclusions, that is, these common pain conditions were more driven by CNS than peripheral mechanisms, and that commonly used analgesics or anti-inflammatory drugs, for example, non-steroidal anti-inflammatory drugs (NSAIDs), opioids and corticosteroids appeared to be ineffective. As such, treatment guidelines for all of these conditions now considered primarily nociplastic all have strongly recommended against the use of opioids for decades, well before there were neurobiological studies showing why opioids might be particularly ineffective in relieving (or even worsen) this type of pain.1 2

Because the protean manifestations of autoimmune and rheumatological disorders involve the entire body and all tissues, rheumatologists are one of just a few specialties that carefully interrogate the entire individual. Not surprisingly then, rheumatologists that played a seminal role in moving fibrositis to FM emphasised the importance of systemic features of FM. Wolfe led the development of all of the FM criteria, and his own work emphasised the importance of using a body map to assess widespreadness of pain, and that FM should be thought of as a continuum (degree of fibromyalgianess) rather than a discrete disorder.3 Wolfe also showed that comorbid FM (what had previously been called secondary FM) was very common in any autoimmune disorder.4 Yunus emphasised the importance of polysensory sensitivity in FM,5 which we now understand to be important in nociplastic pain. Rheumatologists, Don Goldenberg, Buskila and Bennett, emphasised the overlap between FM and other common pain conditions, as well as the familial coaggregation of FM with other pain syndromes.6–8 Hugh Smythe emphasised the finding of diffuse tenderness. Although not a rheumatologist, FM researcher Moldofsky showed the importance of sleep problems in both causing and treating FM.9 Thus, the original 1990 criteria for FM appreciated the presence of diffuse neurally mediated pain sensitivity (originally expressed by assessing the number of tender points) as well as the importance of widespreadness of pain.10

These are all key historical contributions to FM research that later became codified features of what we now call nociplastic pain.11 These coaggregating pain conditions had historically gone by terms such as somatoform disorders, functional somatic syndromes and chronic multisymptom illness, just to name a few.12 As the construct of nociplastic pain has solidified, the preferred terms are now chronic overlapping pain conditions13 (COPCs) or the new International Classification of Diseases (ICD-11) term, primary pain.14 The term primary pain essentially means that the pain is the primary problem rather than being pain secondary to another disease (eg, from osteoarthritis (OA) or rheumatoid arthritis (RA)).

Advances in a number of research techniques, especially functional, chemical and structural brain imaging, helped revolutionise the study of the CNS contributions to pain, and the legitimisation of nociplastic pain. In the first functional MRI (fMRI) study of FM published in 2002, Gracely et al were able to demonstrate that when these individuals were given a low-intensity stimulus that they felt as painful (but a control did not), this led to activation of multiple brain regions known to be involved in pain processing including both the primary and secondary somatosensory cortices.15 This and other early functional imaging studies in conditions such as FM and IBS were beginning to show that these individuals were really experiencing pain, which was a revelation to some who felt that these were mainly neurotic middle-aged women who were feigning symptoms.16 In parallel, quantitative sensory testing (QST) studies showed that in conditions such as FM, individuals were not just tender in regions originally considered as tender points (for this and other reasons leading to the abandonment of tender points from subsequent FM criteria) and instead experiencing diffuse hyperalgesia/allodynia.17 18 These findings strongly suggested a CNS cause of the pain. The term central sensitisation began being used in a different way than the original preclinical studies (that focused on spinal mechanisms19) to indicate that the entire CNS may be sensitised.

One leap in inference came from early QST and functional neuroimaging studies in nociplastic pain states showing that groups of individuals with FM and other nociplastic pain states were just as sensitive (compared with controls) to the sensitivity of light or loudness of noise, as they were to painful stimuli.20–22 This polysensory hyper-responsiveness can only be a CNS phenomenon—visual and auditory stimuli bypass the spinal cord and come directly into the brain via cranial nerves. These findings helped further solidify that nociplastic pain is primarily a CNS-driven disease. Corresponding early brain imaging findings in nociplastic states often identified hyperactivity of brain regions such as the insula, which also made sense in light of revelations by neuroscientists such as Craig and Tracey that were showing that the insular played a key sensory integration role and interoception.23 24

In addition to these studies in primary/nociplastic pain conditions such as FM, similar methods were used to show that nociplastic pain is often present as a superimposed comorbidity in individuals where the primary mechanism of pain is nociceptive, that is, due to ongoing inflammation or damage from autoimmune disorders. For example, early studies showed high rates of FM in nearly every autoimmune disorder, a phenomenon originally termed secondary FM and later central sensitisation or centralised pain.25 In aggregate studies in many conditions were showing that when central sensitisation is present in a nociceptive or neuropathic pain state—that individuals had identifiable differences in pain processing on QST or functional imaging, and became less responsive to a number of peripherally directed therapies.26 For example, in OA, using the 2011 FM Survey criteria as a surrogate measure of central sensitisation/nociplastic pain, the higher the FM score, the less likely individuals respond to lower extremity arthroplasty, as well as opioids used to treat the perioperative pain.27 28 Nearly identical findings were noted in women receiving hysterectomy for chronic pelvic pain.29 30 Soni et al and Tracey et al performed studies in OA showing the fMRI could also predict poor analgesic responses to surgery,31–33 whereas Kosek et al used QST to demonstrate the importance of descending analgesic pathways in predicting response to knee arthroplasty.34 In parallel studies in RA, the FM measure identifies a subset of individuals can be identified that have QST and functional neuroimaging evidence of central sensitisation/nociplastic pain,35–37 and these individuals also appear less likely to respond to biologics and disease modifying anit-rheumatic drugs (DMARDs)38 (see figure 1). The presence of central sensitisation does not simply tell us that peripherally directed therapies will be less effective, this also implies that centrally directed therapies will be more effective. For example, in both OA and RA, serotonin and norepinephrine reuptake inhibitors such as duloxetine and milnacipran are more likely to work in the subset of OA39 and RA40 patients with more widespread pain/nociplastic pain.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The degree of fibromyalgia predicts imaging findings of nociplastic pain in RA (from Nature Review Neurology). DMN, default mode network; FM, fibromyalgia; RA, rheumatoid arthritis.

So it is now clear that (1) nociplastic pain is common both as a sole pain mechanism (eg, FM) or superimposed on nociceptive or neuropathic pain and (2) when an individual has nociplastic pain they will respond less well to peripherally directed therapies and better to centrally directed therapies.

Our current understanding of nociplastic pain

This review is not intended to be an exhaustive review of our current understanding of nociplastic pain. The key findings of a recent review are summarised below41:

1. The best single determinant of pain mechanism is the spatial distribution (widespreadness) of pain.42 The more widespread pain is, and the more pain that occurs where there is no demonstrable inflammation, the more likely that there is a nociplastic component of pain.

2. Symptoms such as fatigue, sleep, memory and mood problems, and sensitivity to non-painful sensory stimuli are also CNS-driven symptoms that are inherent to nociplastic pain. These are not just symptoms that are due to pain—in longitudinal studies, the presence of these other non-pain symptoms predicts the subsequent development of nociplastic pain.43 44 A corollary is that these other symptoms should also be viewed as treatment targets, especially intervening to improve sleep.

3. The risk factors for developing nociplastic pain are becoming increasingly understood and are summarised in figure 2. Females are 1.5–2× more likely to develop nociplastic pain and this difference becomes manifest following puberty. A number of stressors can predispose individuals to be more likely to develop nociplastic pain, but many individuals who are exposed to these same stressors do not develop nociplastic pain. For example, although adverse childhood experiences increase the risk of developing nociplastic pain, only about 10% of those who develop nociplastic pain do so via this pathway.45 Moreover, brain imaging studies suggest that this diathesis to develop nociplastic pain may be identifiable prior to developing pain.46 47

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Risk factors for nociplastic pain (from Nature Reviews Neurology).

1. There are a number of pathophysiological processes occurring in nociplastic pain and these are summarised in figure 3. Functional brain imaging studies show abnormal brain connectivity between brain regions in nociplastic pain. The widespreadness is associated with increased functional connectivity between regions of the default mode, salience and sensorimotor networks in pain conditions with presumed nociplastic mechanisms. In addition, individuals with nociplastic pain have abnormal responses in the descending pain modulatory system, which modulates activity in the spinal dorsal horn and can have inhibitory or facilitatory influences on pain perception. This decrease in inhibitory CNS influences and increase in faciliatory influences leads to a net disinhibition that can be identified by a decreased threshold for nociceptive firing in the periphery.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Risk factors for nociplastic pain across the lifespan (from Nature Reviews Neurology). ACC, anterior cingulate cortex; CSF, cerebrospinal fluid; GABA, gamma aminobutyric acid; IPL, inferior parietal lobule; M1, primary motor cortex; mPFC, medial prefrontal cortext; PCC, posterior cingulate cortex; RVM, rostral ventromedial medulla; S1, primary somatosensory network; SLN, salience network.

There is the activation of the immune system in nociplastic pain conditions. But the inflammation seen in primary nociplastic pain conditions such as FM is quite different than that seen in autoimmune disorders and can be best identified with assays that involve stimulating whole blood to bring out a primed immune system.48 There are also identifiable differences in peripheral nerve fibres, for example, a decrease in intraepidermal nerve fibre density but these are extremely non-specific findings and may likely simply represent neuroplasticity,49 50 akin to the structural changes in the brain that accompany chronic nociplastic pain.51 Nonetheless, the meaning of these changes requires further clarification.

The overall approach to the treatment of nociplastic pain is outlined in box 1.

1. There are (at least) two subsets of nociplastic pain: top-down and bottom-up. Those with top-down nociplastic pain likely have a primary brain disorder that often presents with emerging sleep or memory problems early in life (especially adolescence), followed by pain in one body region, then another and then another as they reach adulthood. The earliest COPCs typically include headache, severe dysmenorrhoea, functional abdominal pain and regional musculoskeletal pains. The familial/genetic contribution to nociplastic pain is highest in these top-down individuals who develop symptoms of primary pain when they are younger.52 53 In contrast, individuals with bottom-up nociplastic pain often have an identifiable nociceptive or neuropathic pain state—and nociplastic pain is superimposed. This had previously been referred to as secondary FM and is now referred to as secondary pain in the current ICD classification. In these individuals, in addition to treating the central contribution to pain, treating the underlying pain condition is often helpful. This may be because this type of central sensitisation is partly driven by ongoing nociceptive input, and/or because improvements in pain often leads to a cascading improvement in sleep and activity and individuals spiral upwards just as they slowly spiralled downwards as they were centralising their pain over many years.54

Where do we go from here?

Do not think of nociplastic pain as being present or absent but rather where an individual sits on a continuum. Wolfe was the first to introduce the concept of fibromyalgianess by showing that in any rheumatic disorder, higher scores on the FM Survey Questionnaire were associated with more pain and disability, and that this measure was often a more powerful predictor of outcomes than more objective measures, for example, ESR, joint space, etc.3 The biggest problem with the FM construct is it forces us to turn a continuum into a binary diagnosis. It is best to consider that some individuals with an autoimmune rheumatic disease will have a bit of super-imposed nociplastic pain, and in some, this will be the primary driver of their pain and other symptoms. Those individuals with mild or moderate nociplastic contributions may respond particularly well to identifying and treating this component of pain since they often have less prominent psychological or functional problems than those with more prominent nociplastic symptoms.

Use a body map (to look for widespread pain) or the FM measure (that also assesses for common comorbid sleep, fatigue, memory and mood problems) in all patients with pain. A body map alone can be an extremely useful tool.42 If individuals are only experiencing pain where there is identifiable damage or inflammation, this is likely to be nociceptive pain. But if the pain is also occurring in regions where there is no identifiable damage or seems to be more intense than one would expect for that amount of damage, then there is likely to be a nociplastic component. When multisite pain is accompanied by other comorbid CNS symptoms seen in nociplastic pain (eg, sleep, memory, mood, sensory hypersensitivity), this makes nociplastic pain even more likely.11

In addition to assessing for the widespreadness of pain, a body map can provide more nuanced clinical information. For example, pain in locations such as the upper arms or upper legs is almost always due to nociplastic pain since there are few diseases or degenerative processes that cause pain in these regions. In each autoimmune disorder, there are expected sites of nociceptive and neuropathic pain, and these sites should not be ‘counted’ as nociplastic when someone has that condition. For example, in RA, we expect to see synovitis and pain in regions such as the hands and wrists, cervical spine, knees and feet, but not regions such as the trunk. We can and should develop disease-specific measures of nociplastic pain in each rheumatic disease that appreciate these differences between rheumatic diseases, to better guide treatment decisions.

The core treatments of nociplastic pain are non-pharmacological. Treatments must be individualised, but the first step is to validate these symptoms and to explain that there are often different underlying mechanisms of pain, that need to be treated in different ways. The diagnosis of nociplastic pain does not in any way imply that peripheral sources of pain do not exist in that individual and may not need treatment. To the contrary, it is important to identify and treat any potential peripheral pain generators. It is better to think of nociplastic pain as an addition element to be treated—rather than the sole pain mechanism.

The core tenets of treating nociplastic pain are to get individuals moving and sleeping. Poor sleep and physical inactivity can directly cause nociplastic pain so focusing first on these comorbidities can be helpful. There are a multitude of ways to get individuals moving and physically active, and the methods that are most accessible to that individual (and if possible those that are patient-chosen) should be tried first. Any slow, gradual improvement in physical activity can be helpful, starting low and going slow. Walking is an excellent exercise, and if individuals cannot tolerate walking then water-based exercise programmes can be extremely helpful. A number of movement-based therapies, for example, yoga, tai chi and Qi Gong, have also been shown to be efficacious. It likely does not matter how you get individuals moving—just that they slowly increase their activity.

Similarly, the methods that can be used to improve sleep range from simple to more intensive. The simplest recommendations are the general sleep hygiene instructions that are readily available on websites. Individuals who do not respond to those simple behavioural adjustments, they may benefit from cognitive–behavioural therapy (CBT) for insomnia, which has been shown to be helpful for many chronic pain conditions. This is also an area where pharmacological approaches can be effective, for example, the use of low doses of tricyclic compounds (eg, cyclobenzaprine, amitriptyline) or gabapentinoids before bedtime.

Analgesics for nociplastic pain. Nociplastic pain appears to be less responsive to both NSAIDs and opioids than nociceptive pain. NSAIDs might still be of benefit for many individuals with nociplastic pain since many have comorbid nociceptive pain, for example, OA that represents a peripheral driver. But opioids should be used cautiously—if ever—in chronic nociplastic pain. There is considerable evidence suggesting that the endogenous opioid system may be involved in the pathogenesis of nociplastic pain states and if that is the case giving someone an opioid could worsen their pain via opioid-induced hyperalgesia.1 2 Supporting this, there is some evidence that blocking the endogenous opioid system with low-dose naltrexone could be of therapeutic benefit in nociplastic pain, although there are other potential mechanisms that could be operative.55

The classes of analgesics that have been shown to be efficacious in several nociplastic pain states include tricyclics, serotonin and norepinephrine reuptake inhibitors (duloxetine, milnacipran, es-reboxetine) and gabapentinoids (pregabalin, gabapentin). All of these drugs have significant off-target effects that limit their usefulness.

Screen for and treat stress, trauma and psychiatric comorbidities. There is a complicated bidirectional relationship between stress and psychological disorders—and nociplastic pain. Trauma and stressors can trigger the development (eg, postinfectious,56 post-war nociplastic states57) or worsening of nociplastic pain. The most recent example is long COVID, where accumulating evidence suggests that nociplastic mechanisms are operative in at least a subset of individuals.58 59

But it is also the case that anxiety, depression and catastrophising that occurs at high rates in individuals with nociplastic pain is partly idiopathic—and related to the fact that these symptoms have often been ignored or maltreated by their healthcare providers for decades. In recently published studies of new-onset nociplastic pain in children, the most common symptom order was that children first developed sleep and memory issues, followed by pain, followed by mood problems.43 In this study, the presence of a mood disorder in a child at baseline did not predict the subsequent development of pain. So at least in young children, the predominant direction is that the pain and resultant functional disability leads to worse mood problems. Similarly, in adults, many recent studies have shown that catastrophising gets markedly better in individuals if you give than effective treatment for their pain.60 61

At the other end of the continuum, there are a significant number of individuals with chronic pain in whom the trauma or psychological stress is clearly driving the pain. These individuals may particularly benefit from recent approaches, for example, emotional awareness and expression therapy (EAET) and/or pain reprocessing theories might be particularly helpful.62–64 Interestingly, in one study of EAET, the individuals who responded best to this approach had relatively normal QST profiles63 suggesting that there are likely many endophenotypes of nociplastic pain. To screen for this type of trauma clinicians can simply ask the patient if they feel that trauma or stress is contributing to (or caused) their symptoms or can use more sophisticated questionnaires that screen for ACEs or a broader set of maltreatments.

We need better care models for the care of this type of pain and these patients since virtually no providers feel comfortable in managing these patients. Rheumatologists are certainly not trained to take care of these types of patients, and in fact, there are very few providers who feel comfortable managing these patients. Even if rheumatologists felt comfortable managing these patients there are far too many of these patients and far too few rheumatologists. Even providers trained in pain medicine do not feel qualified to manage this type of pain since often most of their training is in how to do specific procedures or injections (which will generally not be of benefit).

Although it is a grandiose ‘ask’, this author feels that we need to train a new type of provider with specific expertise in the non-procedural (and non-opioid) management of chronic pain. These could be physicians who are trained as generalists, or advanced care practitioners, who receive additional training to attain this expertise. New care models are also needed that make integrative non-pharmacological more directly available to patients and give patients the option of trying many different types of non-pharmacological approaches. The use of smartphone apps or websites (eg, www.painguide.com) can be very helpful in disseminating such information.

As a field, we should caution against the overinterpretation of findings from animal or in vitro studies suggesting autoimmunity and/or small fibre neuropathy is causing FM and nociplastic pain. As this article outlines, over many decades, the field of rheumatology deduced that individuals with conditions such as FM were quite different than those with autoimmune diseases, causing us to modify the name from fibrositis to FM. Nothing has fundamentally changed. One of the biggest problems in our field is that even though we know have a wide array of DMARDs and biologics, simply treating the inflammatory component of autoimmune disorders does not necessarily make pain and function markedly better.65 We know better than any specialty that nociplastic pain is not responsive to the aggressive anti-inflammatory regimens we typically use to treat autoimmune disorders. This author feels strongly that there is simply no evidence that FM is an autoimmune disease and it is unlikely that any animal or in vitro studies will change that.66 It would require a fundamental redefinition of autoimmunity for that to change since if FM is autoimmune—then so is low back pain and headache—and the term has lost its intended meaning.