Widespread pain: is an improved classification possible?
MacFarlane GJ., Croft PR., Schollum J., Silman AJ.
OBJECTIVE: The classification of widespread pain, proposed by the American College of Rheumatology (ACR) for use in the clinic as a screen for fibromyalgia, as described, does not require truly widespread pain. Studies considering the epidemiology of widespread pain per se may therefore require a definition with greater face validity, which might also show enhanced associations with other physical and psychological measures. We aimed to develop a more coherent definition of widespread pain for use in epidemiological studies and to compare performance in identifying individuals with significant morbidity. METHODS: A group of 172 subjects who had participated in a community based study on the occurrence of pain were identified and categorized by their pain experience as indicated on line drawings of the body according to ACR definition and to a new, more stringent definition that required the presence of more diffuse limb pain. A number of other clinical and psychological measures were recorded for these individuals and the association between their pain status measures and these other variables was assessed and compared. RESULTS: Persons satisfying the newly proposed definition for chronic widespread pain, in comparison with those who satisfied only the present ACR definition, had a significantly higher score on the General Health Questionnaire [median difference (MD) 7.95% CI 1.13], a higher score on the Health and Fatigue Questionnaire (MD 10.95% CI 0.15), and greater problems with sleep (sleep problem score MD 4.95% CI 0.9). Those satisfying the new definition also had a greater number of tender points on examination (MD 3.95% CI -1.7). The morbidity of those satisfying only the present ACR definition was closer to persons who had regional pain. CONCLUSION: A redefinition of widespread pain has produced a group of subjects whose pain is (a) likely to be more "widespread" and (b) is associated more strongly with factors such as psychological disturbance, fatigue, sleep problems, and tender points, and may be more appropriate in epidemiological studies.