Prior to the 1970s and 1980s, fibromyalgia represented an amorphous set of symptoms and definitions—local, regional, or widespread pain with or without other symptoms, including, for some, the idea of “psychogenic rheumatism” (). The American College of Rheumatology (ACR) 1990 criteria de facto established and redefined fibromyalgia to require the presence of widespread pain and multiple (≥11) tender points (). The 2 criteria items were measures of pain, and their diagnostic cutoff levels were determined by comparing patients with fibromyalgia to those without it. But who were the original patients with fibromyalgia and what were they like, this group that constituted the gold standard?
We know now that they were chosen in part by individual physician tolerance of the presence and severity of somatic symptoms, and by the degree of patient distress. The subsequent idea that the 1990 criteria were just about pain was not correct: the criteria implied substantial, if unmeasured, symptomatic distress. In clinical practice after the 1990 criteria were adopted, it was symptoms that made the clinician think of fibromyalgia—and the tender point examination that sometimes followed.
Tender points, the “semi-objective” physical finding, were elusive. Most physicians did not know how to carry out the tender point examination and generally skipped it. When performed, it was most often done incorrectly. The cervical tender points were almost impossible to assess correctly; body habitus and behavioral characters interfered with criteria-provided instructions, and no one really used (or understood) the recommended 4 kg of force. When physicians began the tender point examination, the patient's interview had already provided clues as to what the examination results might be. It is fair to say that the examination may have been influenced by physician beliefs as well as by the contributions of patients. The fibromyalgia examination provided an approximate estimate of tender points—easy at the extremes of tenderness, but much more uncertain in the important middle. Fibromyalgia diagnosis often depended on physician referral, behavioral and emotional characteristics of patients, and the skill, interest, and beliefs of the physicians. Different physicians could and often did come up with different results. In the absence of careful training, the easy assumption that the examination was reliable was usually an untested and dubious assumption.
The 2010 ACR criteria approached the fibromyalgia identification problem differently (). Using the 1990 ACR criteria as the gold standard, fibromyalgia criteria were constructed that excluded tender points, but included a count of pain locations and the physician's rating of the most discriminative symptoms. In these criteria, musculoskeletal pain accounted for about 50% of the criteria score, while the other 50% came from fatigue, sleep, cognitive problems, and an estimate of the overall degree of somatic symptom severity. The relative contributions were not fixed, and it was possible for patients with less musculoskeletal pain to qualify if they had a high enough symptom score. These varying contributions were made explicit in the polysymptomatic distress (PSD) or fibromyalgianess scale (), which combined the musculoskeletal pain and symptom scales by addition.
Similar to the 1990 criteria, the 2010 criteria have practical assessment problems. Physicians are required to interview patients and determine the severity of fatigue, sleep, and cognitive problems, and the overall degree of somatic symptom reporting. It seems certain that physicians will differ in their conscientiousness in making such assessments and their interpretation of the severity of patient complaints.
A self-reported version of the 2010 criteria was developed for research purposes, which should be called the (2011) research criteria. There is strong evidence that it accurately identifies fibromyalgia as diagnosed by clinicians (). Although clinical and epidemiologic results may differ because of sample differences, there is evidence that the 1990 and 2010 criteria have good agreement. But should they really agree completely? The criteria are inherently and deliberately somewhat different because of their differing emphasis on musculoskeletal pain and symptoms. In addition, the 2010 criteria can diagnose patients who might not have sufficient tender points according to the 1990 criteria, but the 2010 criteria also require a sufficiently high symptom score, something that is not a requirement of the 1990 criteria. The differences between individual physicians as examiners and interpreters add additional reasons for differences in criteria results. Still, to imply that the criteria should agree completely suggests that there is really an easily identifiable entity called fibromyalgia. We should not be fooled into thinking that we have a clearly definable entity.
Not a clearly defined entity? So how should we think about fibromyalgia? In the rheumatology literature, fibromyalgia has often been characterized as a specific pain disorder. But it is not seen that way in the broader medical community and literature, where it is looked upon as a type of functional somatic syndrome. Danish investigators characterize fibromyalgia as a bodily distress syndrome, along with chronic fatigue syndrome, irritable bowel syndrome, and similar disorders, because they share a common set of symptoms (). Somatic symptoms may be assessed with the Patient Health Questionnaire-15 (PHQ-15), a well-validated and widely used measure of somatic symptom severity (). Using this scale, Kroenke et al suggested defining illnesses like fibromyalgia as a physical symptom disorder and characterizing symptoms as mild, moderate, or severe. This suggestion “… does not supplant but rather incorporates individual somatic symptoms and syndromes in an umbrella category that recognizes many commonalities.” We recently applied the PHQ-15 to patients satisfying the survey criteria version of the ACR 2010 criteria (). Using a PHQ-15 “moderate” cutoff of 10, a level suggested as consistent with a somatoform disorder (), 89% of 2010 criteria–positive patients had scores ≥10 (moderate or severe somatic symptom severity) ().
The 1990 criteria were 100% composed of signs and symptoms of musculoskeletal pain, the 2010 criteria about 58% of such symptoms, and the PHQ-15 26%. The Pearson's correlation coefficient between the 2010 PSD scale and PHQ-15 was 0.74, but the main difference between the 2010 PSD scale results and those of the PHQ-15 was caused by the increased weight given to musculoskeletal symptoms by the ACR criteria.
Whether we call it fibromyalgia or characterize it as a physical symptom or bodily distress disorder, what we have actually done is take a cloud of different symptoms and symptom severity, including pain symptoms, and organize it into convenient and useful syndromes—but not into discrete diseases or distinctly separate syndromes (). Studies using functional magnetic resonance imaging and similar tools that purport to give insights about fibromyalgia may be tapping into the wider domain of functional and psychological disorders (). Central sensitization, hailed as a biologic marker of fibromyalgia, turned out to be found in almost all painful conditions ([13-15]). Claims that fibromyalgia causes pain or neurobiologic abnormalities need to be examined carefully so that cause and effect are disentangled, and that cause is not confused with mechanism.
Most fibromyalgia patients meet criteria for other functional somatic syndromes and psychological disorders (), and we will find overlaps and comorbidity related to shared genetic and environmental factors. The distress of fibromyalgia symptoms is not dichotomous, but varies in severity, and is found in almost all pain-related disorders. Knowledge of the broad quantity of PSD over its entire severity spectrum, from mild to severe, enlightens patient care and provides a mechanism for assessment and understanding that can be more meaningful and effective than just casting about for a specific diagnosis.