Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.

The Development of Fibromyalgia: Examination of Rates and Predictors in Patients with Rheumatoid Arthritis (RA).

Wolfe1,  Frederick, Hauser3,  Winfried, Hassett4,  Afton L., Katz2,  Robert S., Walitt5,  Brian T.

National Data Bank for Rheumatic Diseases, Wichita, KS
Rush University Medical Center, Chicago, IL
Technische Universität München
University of Michigan Medical School
Washington Hospital Center, Washington, DC


There are no prospective studies of the development of criteria-based fibromyalgia (FM) that consider multiple risk factors. The purpose of the study was to determine rates and predictors of future development of FM in patients with rheumatoid arthritis (RA). We studied FM in RA because of its intrinsic interest and because RA patients provided a substrate of patients that were unselected, particularly with respect to FM characteristics. In addition an extensive set of demographic and severity covariates were available.


After excluding patients with FM and those with high levels of FM symptoms (fibromyalgianess score >10), we studied the development of FM in 9,739 RA patients during 42,591 patient-years of follow-up. We defined FM using a modification of the ACR 2010 FM diagnostic criteria. We used Cox regression to predict future FM, and examined the discriminatory power and accuracy of predictions using Harrell's C concordance coefficient.


At the last observation, 7.4 % of patients satisfied criteria, although 19.8% satisfied criteria at some point during follow-up, for an incidence rate of 5.3 (95% CI 5.1, 5.6) per 100 patients-years, and at rates that were similar in men (7.0%) and women (8.1%). Among those satisfying criteria, during 11,363 years of follow-up from the time of first fibromyalgia diagnosis, approximately half of follow-up time was fibromyalgia+ and was associated with markedly abnormal RA variable and FM variable scores. C-statistics for individual variables, adjusted for age and sex, are shown in the Figure.

In multivariable analyses, demographic factors were weak predictors of fibromyalgia (C=0.604). Stronger predictors were demographic plus RA variables (C= 0.720) and demographic plus fibromyalgia variables (C= 0.765), and all predictors (C= 0.782). Clinically important hazard ratios were noted for cognition, depression, comorbidity and high levels of RA and FM continuous variables. Using a z score of >=5 as a cut point, the most important independent predictors for the multivariable prediction of FM were anti-depressant use, HAQ, fatigue, widespread pain index and the symptom count.


FM can be predicted with clinically available variables. Multiple correlated factors contribute to the development of fibromyalgia. These factors include socio-demographic disadvantage, comorbidity, psychological distress, drug and service utilization, fibromyalgia symptoms (particularly somatic symptom reporting), and functional status; but there is little evidence of the effect of underlying causes. After diagnosis, patients move in both directions across the diagnostic criteria cut point.

To cite this abstract, please use the following information:
Wolfe, Frederick, Hauser, Winfried, Hassett, Afton L., Katz, Robert S., Walitt, Brian T.; The Development of Fibromyalgia: Examination of Rates and Predictors in Patients with Rheumatoid Arthritis (RA). [abstract]. Arthritis Rheum 2010;62 Suppl 10 :112
DOI: 10.1002/art.27881

Abstract Supplement

Meeting Menu