Association of concomitant fibromyalgia with worse disease activity score in 28 joints, health assessment questionnaire, and short form 36 scores in patients with rheumatoid arthritis




To study the association of the presence of fibromyalgia (FM) with the Disease Activity Score in 28 joints (DAS28), the Health Assessment Questionnaire (HAQ), and the Medical Outcomes Study Short Form 36 (SF-36) health survey in patients with rheumatoid arthritis (RA).


A total of 270 outpatients with RA were enrolled in a prospective cross-sectional study. The patients underwent clinical evaluation and application of the HAQ and SF-36 questionnaires. Disease activity was evaluated using the DAS28 score. FM and RA diagnoses were made according to American College of Rheumatology criteria.


The overall prevalence of FM was 13.4%. This group of patients had a higher prevalence of female sex, older mean age, higher functional class, and longer morning stiffness than patients with only RA. Mean ± SD DAS28 scores were significantly higher in patients with RA and FM (5.36 ± 0.99) than in patients with RA only (4.03 ± 1.39; P < 0.001). In a multivariable linear regression analysis, FM was an important predictor of the DAS28 score, even after adjusting for the erythrocyte sedimentation rate, number of swollen joints, functional class, number of disease-modifying antirheumatic drugs currently in use, current dose of steroids, and articular erosions. HAQ and SF-36 scores were also worse in patients with RA and associated FM.


FM is related to worse scores on the DAS28, HAQ, and SF-36 in patients with RA. The presence of FM may have major implications in the interpretation of the DAS28 score because it is related to higher scores independently of objective evidence of RA activity.


Fibromyalgia (FM) has a significant impact on health status, functional capacity, and quality of life (1–5). In patients with rheumatoid arthritis (RA), concomitant FM has been reported in 14–17% of cases (6, 7), and may represent an additional factor that worsens pain and physical, social, and emotional limitations in these patients.

Several instruments are used to evaluate the outcomes in RA. Among them, the Disease Activity Score in 28 joints (DAS28) for disease activity, the Health Assessment Questionnaire (HAQ) for functional status, and the Medical Outcomes Study Short Form 36 (SF-36) health survey for quality of life are the most commonly used (8, 9). There is evidence showing that the presence of FM is associated with a significant increase in the HAQ score in patients with RA (7, 10). One study showed that patients with FM have DAS28 scores similar to patients with RA (11), which is an interesting finding because this instrument was created exclusively for the evaluation of RA (11, 12). Wolfe and Michaud showed lower SF-36 scores in patients with RA and FM when compared with patients with RA only (6). We are not aware of studies evaluating the association of FM with the DAS28 score in patients with RA.

The aim of the present study was to analyze the association of FM with the DAS28 score and to further study the association of FM with worse HAQ and SF-36 scores in patients with RA.



A prospective cross-sectional study was conducted between March 2006 and June 2007. A total of 270 consecutive RA patients attending the outpatient clinic of the Rheumatology Service of the Hospital de Clínicas de Porto Alegre were evaluated. To be included in the study, the patients had to fulfill at least 4 of the 7 classification criteria of the American College of Rheumatology (ACR; formerly the American Rheumatism Association) for the diagnosis of RA (13). Exclusion criteria consisted of refusal to sign the written informed consent, missing results for erythrocyte sedimentation rate (ESR) at the evaluation visit, clinical or laboratory evidence of infection, and overlapping with other connective tissues diseases (except for secondary Sjögren's syndrome). However, patients initially excluded from the study (due to missing laboratory results or active infection) could be included in another visit if no exclusion criteria were present at that time.

Clinical and laboratory evaluations.

The patients were evaluated in 4 sequential stages that occurred during a single visit. First, the SF-36 questionnaire validated for Portuguese (in Brazil) (14) was applied. In the second stage, the same interviewer carried out the specific procedures of the study protocol. Demographic, clinical, and therapeutic aspects of the patients were recorded by interview and chart review. The presence of erosive joint disease was defined according to the evaluation of radiographs of hands and feet performed by an experienced radiologist or rheumatologist. Simple questions (yes or no answers) about symptoms commonly associated with FM were asked, including daytime fatigue (“Do you feel tired during the day, even without physical effort?”), dry mouth (“Do you have the feeling that your mouth is dry?”), dry eyes (“Do you have the feeling that your eyes are dry or that you have sand in them?”), paresthesias (“Do you usually feel tingling, burning, have pins and needles or body numbness?”), headache (“Do you usually have headache?”), mood alterations (“Do you have symptoms of depression or feel anxious, nervous, worried or irritated?”), and nonrefreshing sleep (“Do you wake up tired?”). After that, the evaluation of the presence of diffuse pain according to ACR criteria (15) was made.

In the third stage, the patients completed the HAQ, which was validated for application to Brazilian patients (16). Then a trained doctor, blinded to the SF-36, study protocol, and HAQ data, counted the tender and swollen joints to calculate the DAS28 and recorded his clinical impression of disease activity using a visual analog scale (VAS).

In the final step of the evaluation, the patients were examined for the presence of pain in 18 FM tender points, as recommended by the ACR criteria (15). This examination was performed by a single examiner (AR) who was blinded for the results of all other previous tests. FM was diagnosed according to the fulfillment of ACR criteria (15).

The ESR was measured in the first hour by the Westergren method, using samples collected within 10 days of the evaluation visit. Serum rheumatoid factor was measured by nephelometry, and a value ≥40 IU/ml was considered positive.

This study was approved by the Research Ethics Committee of the Hospital de Clínicas de Porto Alegre, and all patients signed a written informed consent before entering the study.

Calculation of sample size.

Estimating an FM prevalence of ∼17% in patients with RA (6) and a mean ± SD DAS28 score of 4.23 ± 1.5 in patients with RA without FM (11), with a mean difference of 20% between RA and FM patients being considered clinically significant, a sample of 250 patients would have an 89.1% power to detect a significant statistical difference (P ≤ 0.05) in the DAS28 between the groups.

Statistical analysis.

The data were analyzed using Epi Info, version 6 (17) and SPSS for Windows, version 11.0 (18). The association between categorical variables was tested using Pearson's chi-square test, Yates' corrected chi-square test, or Fisher's exact test. Quantitative variables were graphically and statistically tested (with the Kolmogorov-Smirnov goodness-of-fit test) for normality of distribution. Variables with a normal distribution were presented as the mean ± SD, and the between-group comparisons were performed using Student's t-test. Non-normal quantitative variables were presented as the median (25th, 75th percentiles), and the between-group comparisons were performed using the Mann-Whitney test. P values less than or equal to 0.05 were considered statistically significant (all presented P values are 2-tailed).

A multivariable linear regression model was built to evaluate the association of the presence of FM with the DAS28 score, adjusting for confounding variables. The selection of independent variables into the model was based on the capacity of the variable to objectively represent disease activity and/or severity (no automatic method of variable selection was used). Confounding variables that were considered to be significantly influenced by the patient's perception (and therefore directly related to FM itself) were not included. The assumptions of the regression model were assessed by the Kolmogorov-Smirnov test for normality of residuals, White's test for heteroscedasticity, the evaluation of variance inflation factors for detection of colinearity, and tests for nonlinear associations with the aid of the Gretl software (19). Residual analysis was performed to detect Y-dimension outliers, which were defined as cases with studentized deleted residuals with absolute values greater than 3.0. The presence of X-dimension outliers was evaluated by analyzing the weighted leverage values (a value >2p/n was considered high, where n = the number of cases and p = the number of parameters being estimated). Highly influential cases were identified as those with Cook's distances >4/(n − k − 1), where k = the number of independent variables. The logarithmic transformation of independent variables was attempted before exclusion of the case when an extreme outlier was detected. Partial regression coefficients and 95% confidence intervals (95% CIs) were estimated for the independent variables included in the model.


Among 270 patients with RA, 32 (13.4%) fulfilled the criteria for the diagnosis of FM. The demographic and clinical characteristics of patients with and without FM are shown in Table 1. The mean age was higher and women were more prevalent in the group with RA and FM. There was no statistically significant difference between the groups concerning marital status or educational level. Patients with RA and FM had higher functional classes and longer morning stiffness than RA patients, and tended to use prednisone more frequently.

Table 1. Demographic and clinical characteristics of rheumatoid arthritis (RA) patients with and without fibromyalgia (FM)*
 RA (n = 238)RA and FM (n = 32)P
  • *

    Values are the number (percentage) unless otherwise indicated. ES = elementary school; HS = high school; DMARDs = disease-modifying antirheumatic drugs.

  • Pearson's chi-square test, Yates' corrected chi-square test, Fisher's exact test, Student's t-test, or Mann-Whitney test according to the nature and distribution of the data.

  • Cumulative number of DMARDs used since diagnosis.

Age, mean ± SD years55.0 ± 12.460.2 ± 12.80.029
Women197 (82.8)31 (96.9)0.038
Married124 (52.1)12 (37.5)0.173
Educational status  0.830
 Illiterate9 (3.8)2 (6.3) 
 Incomplete ES123 (51.7)18 (56.3) 
 Complete ES/HS/university106 (44.5)12 (37.5) 
Functional class  0.009
 I100 (42.0)4 (12.5) 
 II86 (36.1)20 (62.5) 
 III40 (16.8)6 (18.8) 
 IV12 (5.0)2 (6.3) 
Duration of RA diagnosis, median (25th to 75th percentiles)10.0 (6.0–16.0)10.0 (6.2–13.7)0.713
Duration of RA symptoms, median (25th to 75th percentiles)14.0 (10.0–21.0)14.5 (8.2–22.7)0.998
Number of DMARDs used, median (25th to 75th percentiles)2.0 (2.0–3.0)2.5 (2.0–3.0)0.208
Current use of prednisone126 (52.9)23 (71.9)0.057
Rheumatoid factor positive200 (84.0)26 (81.3)0.884
Presence of joint erosions204 (85.7)23 (71.9)0.080
Morning stiffness for ≥60 minutes76 (31.9)19 (59.4)0.004

Table 2 compares the prevalences of the most common clinical symptoms of FM between the groups. All symptoms were more frequent in the patients with RA and FM. The prevalence of diffuse pain was very low (∼2%) in RA patients without FM.

Table 2. Prevalence of common symptoms of fibromyalgia (FM) in the 2 patient groups*
 RA (n = 238)RA and FM (n = 32)P
  • *

    Values are the number (percentage) unless otherwise indicated. RA = rheumatoid arthritis.

  • Yates' corrected chi-square test, Fisher's exact test, or Mann-Whitney test according to the nature and distribution of the data.

Headache91 (38.2)24 (75.0)< 0.001
Fatigue113 (47.5)28 (87.5)< 0.001
Paresthesias94 (39.5)25 (78.1)< 0.001
Dry eyes100 (42.0)23 (71.9)0.003
Dry mouth108 (45.4)28 (87.5)< 0.001
Sleep disturbance98 (41.2)28 (87.5)< 0.001
Mood disturbance122 (51.3)24 (75.0)0.019
Diffuse pain5 (2.1)32 (100.0)< 0.001
Number of tender points, median (25th to 75th percentiles)4 (1–9)14 (13–16)< 0.001

The DAS28 score was 1.33 (95% CI 0.82, 1.83) points higher in the group with RA and FM when compared with RA patients (Table 3). The difference was related to the subjective components of the DAS28 (tender joints and VAS for disease status), although there was no significant difference in the ESR and the swollen joint count. High disease activity was more prevalent in the group with RA and FM. This group also had a very small prevalence of low disease activity and did not have any patients in remission. The median values of the HAQ, the patient's VAS for pain, and the physician's VAS were also higher in the group with RA and FM.

Table 3. Values of the DAS28 (and related variables), HAQ, patient pain VAS, and physician VAS in rheumatoid arthritis (RA) patients with and without fibromyalgia (FM)*
Evaluation measures of RARA (n = 238)RA and FM (n = 32)P
  • *

    Values are the median (25th to 75th percentiles) unless otherwise indicated. DAS28 = Disease Activity Score in 28 joints; HAQ = Health Assessment Questionnaire; VAS = visual analog scale; ESR = erythrocyte sedimentation rate.

  • Student's t-test, Fisher's exact test, or Mann-Whitney test according to the nature and distribution of the data.

  • Absolute number (percentage).

DAS28, mean ± SD4.03 ± 1.395.36 ± 0.99< 0.001
 ESR, mm/hour25.0 (13.7–40.0)29.0 (16.0–49.0)0.343
 Swollen joints2.0 (0.0–5.0)3.5 (1.0–5.0)0.119
 Tender joints3.0 (0.0–8.0)9.5 (4.5–16.0)< 0.001
 Disease activity VAS32.0 (14.0–53.2)56.5 (42.5–89.5)< 0.001
Disease activity  0.001
 High (DAS28 >5.1)52 (21.8)19 (59.4) 
 Moderate (DAS28 >3.2 to ≤5.1)111 (46.6)12 (37.5) 
 Low (DAS28 ≤3.2)35 (14.7)1 (3.1) 
 Remission (DAS28 <2.6)40 (16.8)0 (0.0) 
HAQ score1.12 (0.62–2.00)2.00 (1.37–2.44)< 0.001
Patient pain VAS40.0 (16.0–66.0)76.0 (52.0–87.2)< 0.001
Physician VAS23.5 (8.7–52.2)53.5 (23.5–67.7)0.001

The values for the SF-36 scales are shown in Table 4. There was a significantly worse quality of life in the group with RA and FM in all aspects except for emotional role, where there was no difference between RA patients with or without FM.

Table 4. Values of the Medical Outcomes Study Short Form 36 health survey scales in patients with isolated rheumatoid arthritis (RA) and in patients with RA and concomitant fibromyalgia (FM)*
 RA (n = 238)RA and FM (n = 32)P
  • *

    Values are the median (25th to 75th percentiles).

  • Mann-Whitney test.

Physical functioning50.0 (20.0–75.0)20.0 (10.0–45.0)< 0.001
Physical role25.0 (0.0–100.0)0.0 (0.0–18.75)0.002
Bodily pain41.0 (31.0–61.0)22.0 (22.0–32.0)< 0.001
General health55.0 (40.0–72.0)47.5 (26.25–52.0)< 0.001
Vitality60.0 (40.0–75.0)30.0 (20.0–50.0)< 0.001
Social functioning75.0 (50.0–100.0)50.0 (25.0–87.5)0.006
Emotional role66.7 (0.0–100.0)66.7 (0.0–100.0)0.829
Mental health68.0 (44.0–84.0)52.0 (21.0–71.0)0.004

A model of multivariable linear regression with the DAS28 score as the dependent variable is shown in Table 5. The results from this model indicated that FM is an independent predictor of the DAS28 associated with a mean adjusted increase of 0.885 points (95% CI 0.551, 1.219) in the DAS28 score. The inclusion of the variables sex and age in the model produced virtually no change in the overall coefficient of determination and partial regression coefficients. When 24 cases identified as outliers or highly influential cases were removed from this model, the multiple coefficient of determination (R2) increased to 0.72 (adjusted R2 = 0.71), and the partial regression coefficient of FM increased to 0.987 (95% CI 0.669, 1.305).

Table 5. Multivariable linear regression model with the Disease Activity Score in 28 joints as the dependent variable*
Independent variablesPartial regression coefficient (95% CI)P
  • *

    Additional results from the multivariable linear regression model: R2 = 0.64; adjusted R2 = 0.63; n = 270. Kolmogorov-Smirnov test of residuals = 0.764, P = 0.604. White's test = 54.461, P = 0.090. 95% CI = 95% confidence interval; ESR = erythrocyte sedimentation rate; DMARDs = disease-modifying antirheumatic drugs.

  • Variables defined numerically as follows: yes = 1, no = 0.

  • The variable ESR was logarithmically transformed to reduce the influence of outliers and improve the multiple coefficient of determination (R2).

  • §

    A quadratic term of the number of swollen joints was suggested during the process of model construction by the nonlinearity test. The variance inflation factors of the variable number of swollen joints and its quadratic term were 6.87 and 7.08, respectively. Other variables presented variance inflation factors <1.15.

  • Constant is the value of the dependent variable when all independent variables are equal to zero.

Fibromyalgia0.885 (0.551, 1.219)< 0.001
Logarithm of the ESR (log10 ESR)1.615 (1.333, 1.896)< 0.001
Number of swollen joints0.344 (0.268, 0.421)< 0.001
(Number of swollen joints)2§−0.014 (−0.020, −0.008)< 0.001
Functional class II or higher0.398 (0.171, 0.626)0.001
Number of DMARDs currently in use0.062 (−0.084, 0.209)0.404
Current dose of prednisone, mg0.014 (−0.004, 0.032)0.131
Articular erosions−0.088 (−0.376, 0.199)0.546
Constant0.789 (0.283, 1.294)0.002


In the present study, we analyzed the association of the coexistence of FM with the results of important instruments of evaluation of patients with RA. Our analyses have shown that FM is significantly associated with worse DAS28, HAQ, and SF-36 scores in patients with RA. Although the DAS28 is one of the main instruments of evaluation of RA activity in clinical trials, we found no previous studies evaluating the association of FM with the results of this scale in patients with RA.

In our sample, 13.4% of patients met the ACR classification criteria for RA and FM simultaneously, which is similar to the previously reported prevalence (6, 7). In contrast to published data, our concomitant FM patients had a higher mean age than those without FM. However, as expected according to FM epidemiology, the group with RA and FM had a greater prevalence of women (6, 7, 10).

Confirming the results of a previous study (7), we observed an association between FM and a higher degree of functional limitation. Patients with RA and FM had a median HAQ score of 2.0 compared with 1.12 for patients with isolated RA; this difference seems relevant because a variation of 0.22 is considered clinically important (20). FM per se has been described as being related to a reduction of the functional capacity similar to that observed in RA (1, 2, 21). Interestingly, FM, which is not usually associated with prolonged morning stiffness, was related to a longer duration of this symptom in patients with RA. Similar results were found by Wolfe et al (10). Perhaps pain, fatigue, and sleep disturbance, which are present in FM, can prolong the morning stiffness sensation typical of RA in patients with both conditions.

Patients with FM and RA had a mean DAS28 score 1.33 (95% CI 0.82, 1.83) points higher than patients with isolated RA. Analyzing the DAS28 components individually, the objective ones (swollen joints and ESR) were not significantly different between the groups. There were also no significant differences in disease duration, number of disease-modifying antirheumatic drugs used, and rheumatoid factor positivity between the groups. Articular erosions tended to be more frequent in patients without FM. However, the ESR was slightly higher and the number of swollen joints tended to be greater in patients with RA and FM. This may indicate that patients with RA and FM actually have a more active disease than patients with RA, but we were not able to (statistically) detect such a difference, possibly due to the size of our sample. Patients with FM also had higher functional class and tended to take prednisone more frequently.

In contrast to the objective components, the subjective components of the DAS28 score (number of tender joints and patient's VAS) showed a great disproportion between the study groups. The small difference in objective components and the large disproportion in the subjective components suggest that the coexistence of FM is related to an increase in the DAS28 score more than would be explained by the disease activity. Therefore, when FM is present, it is probable that the DAS28 does not exclusively reflect the inflammatory activity of RA. This conclusion was corroborated by the results of the multiple linear regression model, which observed an association of FM with higher DAS28 scores independently of several variables that reflect the activity and severity (including the ESR, number of swollen joints, functional status, and prednisone dose) of the disease. This association was maintained at a clinically and statistically significant level (partial regression coefficient 0.668 [95% CI 0.407, 0.907], P < 0.001), even when the physician's VAS of disease activity and duration of morning stiffness were simultaneously added to the model in Table 5.

In the multiple linear regression model, FM was related to a mean increase of ∼1.0 point in the DAS28 score when outliers and highly influential cases were excluded. This increase in the DAS28 score may have an important impact on the classification of the patients. For example, in the group with RA and FM in our study, initially 19 patients (59.4%) had a DAS28 indicative of high disease activity (DAS28 >5.1), whereas 21.8% of patients with isolated RA had high disease activity. Excluding the estimated increase in the DAS28 score related to FM (1.0 point), more than half of these patients (11 of 19) would have been reclassified as having moderate disease activity (DAS28 >3.2 to ≤5.1). Another patient, initially included in the low disease activity group (DAS28 ≤3.2), would have achieved remission criteria for RA (DAS28 <2.6) (9).

The DAS28 score is considered essential to evaluating therapeutic response in clinical trials and is also useful in routine clinical care (22). In addition, the DAS28 is a strong predictor of physical capacity and radiologic progression (23). Therefore, the possibility that FM affects the interpretation of this score may have important implications. In routine clinical practice, misclassification of disease activity may lead to an unnecessary change in the therapy of RA. It may affect the selection of patients in clinical trials, because a high DAS28 score is frequently used as one of the inclusion criteria. The interpretation of results may also be affected, considering that FM symptoms are not expected to respond to therapeutics directed toward RA.

Interesting results were related to the physician's global assessment of disease activity, which was higher in patients with FM and RA. The disproportion between objective evidence of disease activity and the physician's evaluation raises the possibility that the presence of FM affects not only the patient's perception of his/her illness, but may also influence the physician's evaluation.

Studies using the SF-36 and other measures of quality of life have found that FM has a negative impact similar to that of RA (24–27). In the present study, all components of the SF-36 were negatively affected by the presence of FM, except for the impact on psychological aspects in well-being (emotional role), indicating the worse quality of life in this group. These results are similar to those obtained by Wolfe and Michaud (6).

The data from this study demonstrated that the presence of FM is related to higher DAS28 scores, and confirmed the association of FM with worse HAQ and SF-36 scores in patients with RA. The increase in the DAS28 score in patients with RA and FM may lead to significant misclassification of disease activity status. When evaluating and monitoring patients with RA, the presence of associated FM should be considered. In these cases, a more cautious interpretation of the instruments of RA evaluation should be given, and careful clinical judgment may be more appropriate than relying exclusively on numbers when assessing the activity of the disease.


Dr. Ranzolin had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Ranzolin, Brenol, Bredemeier, Feldman, Xavier.

Acquisition of data. Ranzolin, Brenol, Guarienti, Rizzatti, Xavier.

Analysis and interpretation of data. Ranzolin, Brenol, Bredemeier, Xavier.

Manuscript preparation. Ranzolin, Brenol, Bredemeier, Xavier.

Statistical analysis. Ranzolin, Brenol, Bredemeier, Xavier.