The history of the idea of widespread pain and its relation to fibromyalgia

Fibromyalgia – what it is and how it is defined – has been confounded by differing but overlapping clinical and criteria-based descriptions and definitions, all of which have changed over time. Although its antecessors can be seen in the 19th century [1], [2], [3], modern fibromyalgia began in the last half of the 20th century and received meaningful academic and societal recognition in 1990 with the publication of the American College of Rheumatology (ACR) fibromyalgia classification criteria [4].

In the early days of fibromyalgia, which was then known as fibrositis, its musculoskeletal pain component was described in various terms, most of which did not include the concept of all-encompassing or “widespread” pain; rather they spoke of “soreness” and “stiff action” [5], “vague pains shifting from one part of the body to another” [6], “ill-defined musculoskeletal pain, often shifting in location” [7], as well as to the more encompassing “generalized aching and stiffness … they ache all over” [8]. Smythe and Moldofsky in the first, if short-lived, specific criteria for fibrositis required “widespread aching” [9]. The first quantitative but rough definition of fibromyalgia pain came in the widely used 1981 Yunus criteria, “Generalized aches and pains or prominent stiffness, involving three or more anatomic sites (without specifying how many anatomic areas existed) [10].” The Yunus criteria were influential; a Google scholar search indicates 1,190 citations. It was not until the 1990 ACR criteria (cited 9,816 times) that widespread pain was named as such, required a specific assessment of body-wide pain, and became a fibromyalgia criterion. The term, widespread pain, is obviously imprecise, but is usually interpreted to mean something like “distributed over a considerable extent.” But “widespread” could be defined by the extent of body affected (an area measure) and/or specific locations and geometry of affected areas.

In the 1990 fibromyalgia criteria [4], the presence of multiple tender points (tenderness to pressure) was and was intended to be the central criterion; the widespread pain criterion was added to the tender point requirement only so that epidemiological studies could screen patients for fibromyalgia without having to perform the complex tender point examination on every potential subject. The accuracy in classifying fibromyalgia/non-fibromyalgia in the 1990 study was 84.2% using only tender points and 84.8% using tender points and widespread pain. Assessing widespread pain was simple using the 1990 definition, and what was meant to be a technical classification aid grew to be a central part of the fibromyalgia case definition, even though the widespread pain classification accuracy alone was only 65.9% in the ACR study – essentially useless for accurately diagnosing fibromyalgia. After 1990, a series of screening criteria for fibromyalgia also used the ACR definition of widespread pain as an essential component of the fibromyalgia definition. In 2018, additional proposed criteria utilized widespread (or multi-site) pain as the central fibromyalgia criterion [11], and in 2019 the International Classification of Disease (ICD-11) defined fibromyalgia from the perspective of a widespread pain disorder [12]. The practical and research definition of fibromyalgia had changed.

An additional change to the idea of fibromyalgia-linked widespread pain was the transformation of dichotomous widespread pain into a continuous scale as, for example, through the use of the widespread pain index (WPI) [13], [14], [15]. Now, increasing the number of geographic sites of pain could serve not only as a limiting entry to fibromyalgia diagnosis (as in the 1990 criteria) but also as measure of the degree of severity or extent of pain and the specific locations of pain. But what was widespread pain or widespread aching supposed to mean and what was its purpose with respect to fibromyalgia? From the literature and clinical descriptions, it can be seen that the original intent of the widespread pain criterion was to exclude person with geographically limited pain rather than including persons with “true” widespread pain. As an hypothetical example, one might wish to exclude persons who had <30% of body area with pain or aching – an idea that might comport with “ill-defined musculoskeletal pain, often shifting in location” [7]. By contrast, the desire to identify persons with extensive pain or aching, e. g., >80% would be an attempt to identify persons with truly extensive pain and aching, something that was not intended by the clinical or early research criteria for fibromyalgia. In practice, the 1990 ACR definition of widespread pain could be satisfied when as few as three areas in five regions (upper, lower, left, right, axial) were involved, regardless of the size of the regions; as an example, low back, elbow and foot. So, although it was called “widespread pain,” the pain did not have to be widely distributed and often was not; a better term for the 1990 widespread pain definition would have been “more than local or regional pain.”

Over a thirty-year period following the ACR 1990 criteria, the idea of widespread pain became the central part of the fibromyalgia definition. Despite the deficiencies of the ACR definition [16], the idea of widespread pain so defined caught on, and thousands of scientific articles referred to it. To make widespread pain function more satisfactorily in fibromyalgia, however, modern authors de facto altered the definition of widespread pain by introducing the 0–19 widespread pain index (WPI) – a measure of the number of painful sites (or the extensiveness of pain) not just its locations [13], [14], [15], as well as the idea of (extensive) multi-site pain [11]. The principal difference between the older ACR 1990 concept and the modern definitions was that modern definitions required the involvement of more pain sites, effectively more geographic pain. We have shown that the mean WPI increases from 10.4 in those satisfying the 1990 definition to 13.2 using the 2016 fibromyalgia criteria, and 13.0 in the stand alone 2019 widespread pain definition [17].

What is fibromyalgia?

But why should widespread pain be the central part of the fibromyalgia definition at all? That is not what the average clinician thinks of when diagnosing fibromyalgia; [18], [19] nor the conception of the early describers of the syndrome [5], [6], [7], [8], [9]. Only a minority of persons with self-reported reported physician diagnoses of fibromyalgia fulfill fibromyalgia criteria [20], [21]. Non-pain symptoms can be more bothersome than the pain symptoms [22], and persons satisfying fibromyalgia criteria invariably have high levels of distressing somatic and psychiatric symptoms [23], [24]. The 1981 Yunus criteria included weather sensitivity, aggravation of symptoms by anxiety or stress, poor sleep, general fatigue or tiredness, anxiety, chronic headache, irritable bowel syndrome, and numbness [10]. The tender point count of the 1990 criteria was strongly correlated with somatic symptoms, even though the 1990 criteria did not contain any somatic symptoms scales [25], [26], [27]; and the 2010–2016 ACR and modified ACR fibromyalgia criteria give almost equal weight to widespread pain and somatic and psychiatric symptoms via the WPI and SSS scales. At the ACR annual meeting following the publication of the 1990 criteria, Xavier Caro MD reported from the podium that for each of a long list of clinical items examined, patients with fibromyalgia had more and more severe symptoms compared with those without fibromyalgia. “Is there nothing,” a questioner in the audience asked, “that is not more common or more severe in fibromyalgia?”

We would make the case that fibromyalgia is really about multiplicity and severity of pain and somatic and psychiatric symptoms, and that separation of patients according to the degree that they fit a widespread pain definition – while convenient for research and clinical definition – represents a logically false classification, a step that can be avoided through the use of the polysymptomatic distress (PSD) scale of the ACR and modified ACR criteria [13], [14], [15].

Where does widespread pain fit in and how does it work?

It is that pain extent and somatic and psychiatric symptoms are all highly correlated. The more pain, the more non-pain symptoms. The more non-pain symptoms, the more pain symptoms. This relationship holds both below and above the fibromyalgia threshold (Figure 1). In our studies, the correlation between the polysymptomatic distress scale (PSD), which is a continuous scale estimator of fibromyalgia, and the widely used Somatic Symptom Scale-8 (SSS-8) [28] is 0.750 (unpublished data). This relationship is the same whether the SSS-8 scale or the ACR Symptom Severity Scale (SSS) is used (Figure 1, upper left and right), and it also exists when non-pain, non-fibromyalgia symptoms (NPNS) are visualized (Figure 1, lower left). It can be seen that cut point between fibromyalgia and non- fibromyalgia shown by the vertical line in the figure is arbitrary, and by altering the required level of widespread pain, the definition of fibromyalgia can be altered. As there is no gold standard, the definition of fibromyalgia is always arguable and open to question. Different definition of widespread pain will result in different patients being identified as satisfying fibromyalgia criteria [29]. Perhaps more importantly to patient care and understanding symptoms, the relation of WPI to symptoms is present through the full range of fibromyalgia and non-fibromyalgia levels. This relationship should be thought of as natural and expected. The answer to the question “Is there nothing that is not more common or more severe in fibromyalgia” is no. As long as fibromyalgia is defined as the end of a spectrum of pain and symptom severity, symptoms will always be more common and more severe in persons in that region. As discussed above and shown in Figure 1, fibromyalgia, PSD, WPI, SSS-8 are all highly correlated. It might be a step forward to loosen the dependence of fibromyalgia on widespread pain and think of fibromyalgia in terms of a bodily distress or functional somatic syndrome [30], [31].

Figure 1:

The distribution of widespread pain, the relationship between the widespread pain index (WPI) and the Somatic Symptom Scale-8 (SSS-8), the ACR symptom severity scale (ACR SSS), and a 0-24 index of non-pain-non fibromyalgia symptoms (NPNS). Data are from the study of widespread pain in this journal [17].