In this issue of Arthritis Care & Research, Ranzolin and colleagues warn of the danger of accepting Disease Activity Score (DAS) scores as “truth” when rheumatoid arthritis (RA) patients have fibromyalgia (1). Information like this should make us think about the validity of the 28-Joint Count DAS (DAS28) but also about the validity and meaning of fibromyalgia.
Let's look at fibromyalgia first. Fibromyalgia is a clinically convenient term to identify patients with high levels of polysymptomatic distress (fatigue, somatic symptoms, cognitive and sleep problems, decreased pain threshold, among many additional problems) (2). Pain threshold can be sensed by the number of painful nonarticular regions that patients report (3). We can call polysymptomatic distress “fibromyalgianess,” and we can model it with the Symptom Intensity (SI) scale, a measure that combines the number of nonarticular painful sites and a fatigue visual analog scale (4). Figure 1 shows the SI scale in 23,622 RA patients from the National Data Bank for Rheumatic Diseases.The mean SI scale score of patients with fibromyalgia is 7.25 (approximately the 92nd percentile), while the mean SI scale score for all RA patents is 3.75.
All patients, indeed all of us, can be found on the scale shown in Figure 1, for the tendency to respond with distress to physical and mental stressors is part of the human condition. It is clear from the Figure that patients with SI scores of 5 and 6 will also have distorted DAS scores, even though they may not have fibromyalgia. This figure also illustrates the problem and fallacy of categorizing an “ends of spectrum” (5) disorder as a disease; and it illustrates a major validity problem with fibromyalgia research where fibromyalgia patients (at the right tail in Figure 1) are compared with “controls” (on the left in Figure 1). The neurobiologic and imaging abnormalities found in fibromyalgia (6) and the consequent (erroneous) idea that fibromyalgia is a disease are a function of sampling and selection bias. It is easy to demonstrate statistical significance when you are examining the ends of a distribution. It is more difficult, but much more informative to examine the entire distribution. Even simple dichotomization loses information and distorts meaning (7). Fibromyalgianess is real, and worthy of study, but it must be studied correctly.
With respect to DAS28 problems, it is worth noting that there are also patients with very low SI scores. RA patients with high tolerance to pain and RA damage (“typus robustus”) were described more than 35 years ago (8). If fibromyalgia patients have an inappropriately high DAS score, then robust patients and those with low SI scores will have inappropriately low scores. Because DAS scores utilize the patient's global and the patient's report of tender joints, variables that are strongly related to fibromyalgianess, they will very often be somewhat too high or too low compared with the average patient.
All of this means that we should look at patients and the individual components of the DAS28, not just at the index score. Indeed, we should consider using a visual analog fatigue scale to tell us more about fibromyalgianess. What we should not do is to mindlessly believe the DAS28 score, and we and our professional organizations should resist demands from regulatory authorities and insurance companies to do that (9). Throw the mountains into the lakes and Switzerland, on average, is flat.