• Fibromyalgia syndrome;
  • Operant conditioning;
  • Cognitive–behavioral therapy;
  • Treatment


  1. Top of page
  2. Abstract


To predict the effects of cognitive–behavioral therapy (CBT) and operant–behavioral therapy (OBT) in fibromyalgia syndrome (FMS).


A total of 125 patients who fulfilled the American College of Rheumatology FMS criteria were randomly assigned to CBT (n = 42), OBT (n = 43), or attention placebo (AP; n = 40). The pretreatment to 12-month followup reliability change index was used to determine clinically meaningful changes in pain intensity and physical impairment. Multinominal logistic regression analyses were used to determine the predictors of improvement in pain intensity and physical impairment for the entire sample. Analyses of variance were computed to compare the characteristics of responders and nonresponders in each of the 3 interventions.


At the 12-month followup, 53.5%, 45.2%, and 5% of patients in the OBT, CBT, and AP groups, respectively, reported clinically meaningful improvements in pain intensity. Similarly, 58.1%, 38.1%, and 7.5% of patients treated with OBT, CBT, and AP, respectively, reported clinically significant improvements in physical impairment. Prior to treatment, the OBT physical impairment responders displayed significantly more pain behaviors, physical impairment, physician visits, solicitous spouse behaviors, and level of catastrophizing compared with nonresponders. The CBT physical impairment responders, compared with nonresponders, reported higher levels of affective distress, lower coping, less solicitous spouse behavior, and lower pain behaviors.


The results of this study suggest that pretreatment patient characteristics are important predictors of treatment response and may serve as a basis for matching treatments to patient characteristics. Prospective outcome studies are needed to confirm whether the tailoring of treatment actually leads to better outcomes for patients with FMS.


  1. Top of page
  2. Abstract

Fibromyalgia syndrome (FMS) is defined by widespread pain (at least 3 quadrants of the body and axial) for at least 3 months, and tenderness at 11 or more of 18 tender points located throughout the body (1). Patients diagnosed with FMS also report poor coping ability (2, 3) and display overt expressions of pain (i.e., pain behaviors [4, 5]), suggesting that psychological processes may contribute to the maintenance of symptoms and disability (6, 7).

The efficacy of a diverse set of interventions has been inconsistent. Moreover, even when the results of treatment are statistically significant, a large proportion of patients do not report clinically significant outcomes. Thus, despite many clinical trials, there is no consensus regarding optimal management of FMS. A clinical practice guideline has been proposed (8) emphasizing education, exercise, antidepressant medication, and cognitive–behavioral therapy (CBT).

Multidisciplinary treatments for FMS with a cognitive–behavioral orientation result in significant changes in pain-related convictions of control, reductions in pain intensity and interference with life, and improvements in emotional distress (9, 10). In a previous study, a clinically significant reduction in pain severity measured with the reliability change index (11) was found in 42% of the sample and was maintained for 6 months (10). One recent meta-analysis of CBT in patients with FMS, however, questioned the efficacy of CBT (12).

Operant–behavioral therapy (OBT), based on the principles of operant conditioning and the importance of the antecedents and consequences of behavior (13), has been reported to produce a significant and stable reduction in pain intensity, interference, pain behaviors, health care utilization, and improvement in sleeping for patients with FMS (14–16). Sixty-five percent of the patients treated with OBT reported clinically significant improvement compared with none of the patients who received physical therapy alone (15). Although the results reported with OBT are promising, there have only been a few studies that have evaluated the effectiveness of this treatment in patients with FMS.

A previous comparison of CBT and OBT revealed that CBT influenced pain-related, cognitive, and affective variables, whereas OBT showed improvements in functioning and behavioral variables in addition to pain intensity (15, 17). Therefore, these treatments appear to have an impact on different variables. Even though the results that have been reported indicate statistically significant improvements on some outcome criteria, a significant percentage of patients do not demonstrate clinically significant benefits (17). Most of the published clinical outcome studies have treated patients as homogeneous groups based on the diagnosis of FMS despite the patient heterogeneity observed in both physiologic variables (18–20) and psychosocial characteristics (2, 21, 22). The results observed with CBT and OBT raise an important question regarding which patients are most likely to benefit from these treatments, which have different emphases.

Several articles (3, 10, 23, 24) report predictors for treatment efficacy in patients with chronic back pain, temporomandibular pain, FMS, and rheumatoid arthritis following various psychological treatments. Although the efficacy of electromyogram biofeedback was predicted by chronicity and treatment-specific variables such as low reactivity of the affected muscles, the efficacy of CBT was predicted by cognitions and coping resources (3, 10, 24), duration of the disorders (24), depression, and low solicitous behavior (10).

This report contains a secondary analysis of the results of a previous study (17) comparing the effectiveness of CBT with OBT and with an attention placebo control (AP) in patients with FMS. The results of the previous study indicated that although both treatments provided evidence supporting the efficacy of the 2 treatments compared with a control group, not all patients achieved the same positive outcomes. The primary goal of the present study was to identify the characteristics of CBT and OBT responders and to compare them with each other and with an AP group. Identification of responder criteria could serve as the basis for prospective treatment to determine whether prescribing treatments based on identified responder criteria would produce improved outcomes compared with offering generic treatments to all patients with FMS.


  1. Top of page
  2. Abstract


A sample of 125 consecutive married women diagnosed with FMS were recruited from outpatient rheumatologic clinics. The mean ± SD age of the participants was 46.71 ± 10.58 years (range 21–67 years), mean duration of symptoms was 10.53 ± 9.74 years (range 0.6–33 years), a mean of 6.66 ± 2.56 regions (range 3–10) were reported to be painful, and the mean number of positive tender points (Manual Tender Point Survey [MTPS] [25]) was 15.01 ± 1.26 (range 11–18) with a mean pain intensity of 4.95 ± 2.28 (range 1–9.52) on a scale ranging from 0 (not at all painful) to 10 (extremely painful) (Table 1). The treatment and control groups had comparable demographic factors (Table 1).

Table 1. Demographic and clinical data of the patients (n = 125)*
 OBT (n = 43)CBT (n = 42)AP (n = 40)
  • *

    Values are the mean ± SD (range) unless otherwise indicated. OBT = operant–behavioral therapy; CBT = cognitive–behavioral therapy; AP = attention placebo.

Age, years43.23 ± 9.0349.13 ± 10.0347.46 ± 9.75
Duration of pain, years8.98 ± 10.119.08 ± 8.508.73 ± 8.77
No. painful regions7.41 ± 1.816.74 ± 2.187.06 ± 2.07
No. tender points16.66 ± 3.9317.25 ± 4.4716.88 ± 4.33
Tender point pain intensity5.73 ± 2.144.59 ± 1.494.21 ± 1.73
No. physician visits36.87 ± 15.1530.55 ± 16.2034.25 ± 16.33
Drug (no./day)3.62 ± 2.163.16 ± 3.623.36 ± 2.78
Occupational status, no. (%)   
 Working17 (39.5)19 (45.2)20 (50.0)
 Unemployed16 (37.2)16 (38.1)12 (30.0)
 Workers' compensation3 (7.0)2 (4.8)3 (7.5)
 Retired7 (16.3)4 (9.5)3 (7.5)
 Student0 (0.0)1 (2.4)2 (5.0)

Treatment protocol.

The patients were randomly assigned to either the OBT, CBT, or AP group. The study adhered to the guidelines of the Declaration of Helsinki, the local institutional review board approved the study, and informed consent was obtained from all participants.

All patients underwent medical and rheumatologic assessments, and all fulfilled the American College of Rheumatology criteria for FMS (1). The inclusion criteria consisted of 1) pain for at least 6 months, 2) married, 3) willingness of the spouse to be involved, and 4) fluency in German. There were no significant differences between the patients assigned to the 3 groups with respect to age, duration of FMS, pain intensity, or number of painful areas (Table 1). The exclusion criteria consisted of inflammatory rheumatologic diseases and any concurrent major disease such as cancer, diabetes, or kidney failure.


The assessments were described in detail in a previous article (17). In summary, they consisted of assessment of blood chemistry, neurologic examination, and evaluation of number of painful points and pain intensity on palpation of the FMS tender points using the MTPS (25). Participants also completed the Fibromyalgia Impact Questionnaire (FIQ) (26, 27), the Multidimensional Pain Inventory (MPI) (28, 29), and the Pain-Related Self-Statements Scale (PRSS) (30). In addition, pain behaviors were elicited by asking participants to engage in a series of daily activities (e.g., window washing). Patients were videotaped performing these tasks (5) and pain behaviors were assessed with the Tübingen Pain Behavior Scale (TBS) (31). The TBS rates the presence of behaviors on a 0–2 scale (where 0 = none, 1 = sometimes, and 2 = always). Coding was carried out separately for the conditions “presence of spouse” and “absence of spouse.” All patients completed treatment expectation ratings before the first session and satisfaction ratings at the end of the first and last sessions. Satisfaction was rated on a 6-point scale ranging from 0 (completely unsatisfied) to 6 (completely satisfied). This measure was included as a means of determining whether the groups differed in their beliefs about the quality of the treatment received.

Medication consumption and the number of physician visits in the 12 months prior to and after treatment were extracted from the medical records. At the followup, these visits were obtained from the patients' medical records that were maintained at the clinics the patients were scheduled to attend at 6-week intervals.


Each treatment consisted of 15 weekly 2-hour sessions that were co-led by a psychologist and a rheumatologist and were conducted in groups of 5 patients. Spouses attended sessions 1, 5, 9, and 13.

Cognitive–behavioral therapy.

CBT focused on the patients' thinking and involved problem solving, pain coping strategies, and relaxation (32, 33). Patients and spouses received weekly homework tasks, were encouraged to engage in physical activities, and were asked to reduce analgesic medication gradually over the course of treatment. Patients participated in relaxation exercises during and between the sessions.

Operant–behavioral therapy.

OBT was based on changing observable pain behaviors and included video feedback of expressions of pain, contingent positive reinforcement of pain-incompatible behaviors, and punishment of pain behaviors. The treatment also included time-contingent intake and reduction of medication, increase of activity, reduction of interference of pain with activities, reduction of pain behaviors, and training in assertive behaviors (32). In contrast to CBT, OBT focused on behavioral expressions of pain without directly targeting maladaptive thoughts. Patients engaged in role plays to reduce pain behaviors and increase healthy behaviors. Patients, their spouses, and group members used a reinforcer plan that consisted of the presentation of a red card when pain behaviors were displayed and a green card when healthy behaviors were displayed. Patients were encouraged to increase their activity levels and were assigned homework to increase activities and reduce pain behaviors. Medication was reduced immediately following the assessment.

Attention placebo.

AP focused on general, therapist-guided discussions. The discussions were centered around medical and psychosocial problems of FMS (i.e., stress in different areas of patients' lives and use of medication). Therapists did not initiate topics and made no specific recommendations. Patients did not receive any specific homework.

Statistical analysis.

The intent-to-treat principle guided the analyses. Baseline scores for patients who dropped out of treatment were carried forward. The primary outcome measures were changes in pain intensity (28, 29), physical functioning (26, 27), and affective distress (28, 29) at posttreatment, 6-month followup, and 12-month followup.

Between-group effects.

The results of the initial study (17) revealed that patients in both the OBT and CBT groups reported statistically significant reductions in pain intensity posttreatment. In addition, the CBT group reported statistically significant changes in cognitive and affective variables and the OBT group demonstrated statistically significant improvements in functioning and behavioral variables. The AP group reported no significant improvements in any of the outcome variables. The posttreatment effects for the OBT and CBT groups were maintained at 6-month and 12-month followup. These results suggest that both OBT and CBT are effective in treating patients with FMS, with some differences in the outcome measures targeted by the individual treatments compared with an unstructured social discussion. In the AP group, deterioration in symptoms and functioning was observed.

Clinical significance.

The reliability change index (RC) (11) was computed as RC = X2 − X1/S1equation image, where X1 = T1-value and X2 = T4-value, S1 = standard deviation, and Rxx = test–retest reliability. The RC index is a conservative measure used to determine the clinical meaningfulness of change used as the basis for determining the responder rate to treatment. The pain severity scale of the MPI (28) and the physical impairment scale of the FIQ (26) were used to define clinically meaningful changes.

Responder criteria.

All treatments shared several characteristics, namely, attention and group experience. Therefore, we attempted to identify predictors of response in the groups combined. We used the RC index to identify responders with the outcomes pain severity (MPI) and physical functioning (FIQ). Responders were prospectively defined as patients showing a ≥50% reduction in pain score and physical impairment. Multinomial logistic regression analyses were performed with responders with clinically significant improvement of pain and physical impairment, responders with clinically significant deterioration of pain and physical impairment (negative responders), and nonresponders with no change in pain and physical impairment. Univariate comparisons were initially conducted and then variables with a P value less than 0.05 (5% significance level) were included in the multivariate analyses. The predictor variables were entered in the following order as sets: baseline pain intensity (MPI), physical impairment (FIQ), affective distress (MPI), coping and catastrophizing (PRSS), behavioral variables (solicitous spouse responses in the MPI, pain behaviors, number of physician visits), and duration of pain.

To assess predictors of treatment response as 1 outcome, the initial values of the physical functioning of responders and nonresponders were compared. These calculations were followed by one-way analyses of variance (ANOVAs) to determine differences between responders with clinically significant improvement of physical impairment and nonresponders of each treatment group to identify the responder criteria for the CBT and OBT. A canonical discriminant analysis was performed that included demographic, baseline pain severity, physical functioning, affective, cognitive, and operant variables to determine the relevant discriminant variables that would significantly differentiate groups.


  1. Top of page
  2. Abstract


Three patients in the OBT group (6.9%), 2 in the CBT group (4.8%), and 20 in the AP group (50%) stopped prematurely. The primary reason that patients gave for dropping out of the AP group was deterioration of symptoms. Patients who stopped prematurely were not significantly different from those who completed treatment in duration of symptoms, initial pain severity, or number or severity of tender points. Overall, 100 patients completed the treatments, 40 in the OBT group, 40 in the CBT group, and 20 in the AP group.

Treatment expectation and satisfaction.

There were no statistically significant between-group differences in treatment expectation (F[2,122] = 1.47, P = 0.24) in the first treatment session. For treatment satisfaction (first and last session), an ANOVA revealed neither a significant group effect (F[2,122] = 1.42, P = 0.25) nor a significant group × phase effect (first versus last session; F[2,122] = 0.53, P = 0.59).

Clinical significance of the improved and deteriorated changes.

The RC index of the MPI pain scale showed a responder rate of 53.5% for the OBT group, 45.2% for the CBT group, and 5% for the AP group (χ2[2] = 31.51, P < 0.001) 12 months following treatment. The percentage of CBT and OBT responders was significantly higher than that of the AP responders (CBT versus AP: χ2[1] = 16.30, P < 0.001; OBT versus AP: χ2[1] = 21.36, P < 0.001). CBT and OBT responders did not differ significantly (Figure 1). Clinically significant deterioration was observed only in the AP group (40%; χ2[2] = 35.23, P < 0.001).

thumbnail image

Figure 1. Responder rate related to clinically significant reduction or increase of pain intensity 12 months after cognitive–behavioral therapy (CBT), operant–behavioral therapy (OBT), and attention placebo (AP).

Download figure to PowerPoint

The RC index of the FIQ physical impairment scale indicated responder rates of 58.1%, 38.1%, and 2.5% in the OBT, CBT, and AP groups, respectively (χ2[2] = 14.95, P < 0.005). The percentage of CBT and OBT responders was significantly higher than that of the AP responders (CBT versus AP: χ2[1] = 7.23, P < 0.05; OBT versus AP: χ2[1] = 14.02, P < 0.001). CBT and OBT responders were not significantly different (Figure 2). Clinically significant deterioration was observed in 6.9%, 14.3%, and 47.5% of the OBT, CBT, and AP groups, respectively (χ2[2] = 5.63, P < 0.05).

thumbnail image

Figure 2. Responder rate related to clinically significant reduction or increase of physical impairment (PhI) 12 months after cognitive–behavioral therapy (CBT), operant–behavioral therapy (OBT), and attention placebo (AP).

Download figure to PowerPoint

Responder criteria for change in pain intensity in the entire sample.

Reduction of pain intensity.

In the first multinominal logistic regression, lower initial physical impairment (B = −0.699, SE = 0.245, P = 0.004, exp[B] = 0.497) was associated with a clinically significant decrease in pain intensity for treatments combined. In comparison with nonresponders, high physical impairment prior to treatment decreased the likelihood of clinically significant improvement of pain ∼0.5 times. Low physical impairment prior to treatment predicted a clinically significant pain reduction (Table 2). Physical impairment (χ2[2] = 11.25, P = 0.025) accounted for 27% of the reduction of pain intensity for the entire sample following psychological pain treatment. Duration of pain, psychological factors, and behavioral factors did not predict reductions in pain intensity. There were no significant predictors for responders with clinically significant deterioration of pain for all treatments combined.

Table 2. Adjusted odds ratios for responders with clinically significant reduction of pain/physical impairment and responders with clinically significant increase of pain/physical impairment*
Predictor variablesResponders with clinically significant reductionResponders with clinically significant increase
Pain intensity (n = 44)Physical impairment (n = 44)Pain intensity (n = 16)Physical impairment (n = 33)
  • *

    Values are the odds ratio (95% confidence interval).

Physical impairment0.49 (0.31–0.80)4.24 (1.29–13.87)11
Pain intensity 1 0.02 (0.00–0.35)
Affective distress 0.11 (0.02–0.71) 11.63 (0.92–14.07)
Solicitous spouse behavior 0.23 (0.07–0.80) 1
Pain behavior 1.05 (1.00–1.09) 1
Number of physician visits 1 1.12 (0.99–1.26)

Responder criteria for reduction and increase of physical impairment in the entire sample.

Reduction of physical impairment.

In the second multinominal logistic regression, higher initial physical impairment (B = 1.444, SE = 0.605, P = 0.007, exp[B] = 4.237), lower initial affective distress (B = −2.241, SE = 0.966, P = 0.010, exp[B] = 0.106), and solicitous spouse behavior (B = −1.466, SE = 0.632, P = 0.007, exp[B] = 0.231), as well as higher initial pain behaviors (B = 0.047, SE = 0.022, P = 0.009, exp[B] = 1.048) were predictive of clinically significant decreases in physical impairment. In comparison with nonresponders, greater physical impairment prior to treatment increased the likelihood of clinically significant improvement of physical impairment ∼4.24 times, high affective distress and greater solicitous spouse behavior prior to treatment decreased the likelihood of clinically significant improvement of physical impairment ∼0.11 and 0.23 times, respectively, and greater pain behavior prior to treatment increased the likelihood of clinically significant improvement of physical impairment ∼1.05 times. Greater physical impairment and pain behavior and lower affective distress and reduced solicitous spouse behavior predicted a clinically significant reduction in physical impairment (Table 2).

Increase in physical impairment.

Lower initial pain intensity (B = −3.776, SE = 1.385, P = 0.006, exp[B] = 0.233), higher initial affective distress (B = 2.453, SE = 1.295, P = 0.005, exp[B] = 11.625), and a larger number of physician visits (B = 0.113, SE = 0.062, P = 0.048, exp[B] = 1.119) were significantly associated with a clinically significant increase in physical impairment. In comparison with nonresponders, high pain intensity prior to treatment decreased the likelihood of clinically significant deterioration of physical impairment ∼0.23 times, high affective distress and a large number of physician visits increased the likelihood ∼11.63 times and ∼1.12 times, respectively. High affective distress in connection with low pain intensity and enhanced pain behavior predicted clinically significant deterioration of physical impairment (Table 2).

The combination of pain (χ2[2] = 13.92, P = 0.001), physical variables (χ2[2] = 14.91, P = 0.001), affective variables (χ2[2] = 20.74, P < 0.001), and behavioral variables (χ2[2] = 14.97, P = 0.001) explained a total of 65.9% of the variance of physical impairment following treatment. Duration of pain and cognitive factors did not predict significant reductions in physical impairment.

Canonical discriminant analyses found that pain behavior discriminated the CBT group from the OBT and AP groups (Wilks' λ[2] = 0.488, P < 0.001; F[2] = 11.561, P < 0.001). CBT patients responding with clinically significant reduction of physical impairment showed lower pain behaviors prior to the treatment in contrast to OBT responders who showed a greater number of pain behaviors and AP patients who displayed extremely high pain behaviors.

Responder analyses for individual treatments.

Prior to treatment, responders to CBT showed significantly lower levels of affective distress (F[1,32] = 4.21, P < 0.005), lower coping (F[1,32] = 8.47, P < 0.005), fewer solicitous spouse behaviors (F[1,32] = 14.78, P < 0.001), and lower pain behaviors (F[1,32] = 12.23, P < 0.001) compared with the nonresponders. Prior to treatment, responders to the OBT treatment reported significantly higher levels of physical impairment (F[1,35] = 5.07, P < 0.01) and pain behaviors (F[1,35] = 13.66, P < 0.001), a greater number of physician visits (F[1,35] = 6.03, P < 0.001), more solicitous spouse behaviors (F[1,35] = 7.31, P < 0.005), and a higher level of catastrophizing (F[1,35] = 5.21, P < 0.01) compared with the nonresponders.

The only variable that differentiated between AP responders and nonresponders was spouse response. Prior to treatment, responders in the AP group reported higher negative spouse behavior (F[1,18] = 8.51, P < 0.01) compared with the nonresponders. Because a high number of participants in the AP group experienced deterioration, variables that predicted deterioration were examined. Specifically, patients with extremely high pain behaviors (F[1,18] = 5.59, P < 0.05) and high physical impairment (F[1,18] = 5.14, P < 0.05) showed a clinically significant deterioration after AP.


  1. Top of page
  2. Abstract

Examination of the responders revealed that 45% of the patients treated with CBT demonstrated clinically significant reduction in pain maintained over 12 months following treatment. The responder rate was comparable with that in the study by Turk et al (10). Consistent with previous studies (9, 10), successfully treated patients reported reduced pain intensity, less affective distress, more coping, and less catastrophizing. OBT responders (54%) showed reduced physical impairment, fewer physician visits, and reduced pain behaviors.

The results obtained are also consistent with other studies of patients with FMS treated with psychological approaches (15, 16). The CBT and OBT groups had larger clinically significant reductions in physical impairment compared with the AP group. In contrast to 7.5% of the patients in the AP group, 58% of the patients treated with OBT and 38% of the patients treated with CBT significantly reduced their physical impairment.

The clinically significant deterioration of 47.5% of the patients in the AP group was related to extremely high pain behaviors and physical impairment. These variables suggest that the social discussion (AP) may have inadvertently reinforced pain and disability in a group that was already prone to operant reinforcement of pain. The pain behaviors and physical impairment of the CBT and OBT negative responders (6.9% and 14.3%, respectively) were not significantly different compared with the AP negative responders. In contrast, CBT and OBT responders with clinically significant improvements showed significantly lower pain behaviors and physical impairment than the negative responders. All negative responders showed high pain behaviors and levels of physical impairment. That is, patients with extreme values did not profit from any psychological treatment.

Pain responders to a psychological treatment had lower pretreatment physical impairment in accordance with Williams (34). Psychological variables did not predict significant reductions in pain. This is not surprising because pain reduction is not the primary end point for psychological treatments. Rather, improvement in physical functioning is typically the objective.

High levels of physical impairment and pain behavior and lower affective distress and reduced solicitous spouse behavior predicted improvement in physical functioning. Consistent with these results, Turk and colleagues (10) found that treatment failure was associated with a high level of emotional distress, high perceived disability, and low perceived life control.

Several variables differentiated patients who benefited from CBT from those who improved with OBT. Whereas patients with more pain behaviors and physical impairment, a higher number of physician visits, more solicitous spouse behaviors, and a higher level of catastrophizing benefited more from OBT, patients with lower pain behaviors, higher levels of affective distress, lower coping, and reduced solicitous spouse behaviors appeared to obtain the greatest benefits from CBT. These results are consistent with other studies (3, 23, 24) that found that coping, cognitive factors, and seeking support from others accounted for 18% of the variance in improvement of pain-related interference (3, 10).

Pain behaviors were the most important predictor that discriminated the efficacy of CBT in comparison with OBT and AP. Whereas CBT responders showed low pain behaviors, OBT responders displayed high pain behaviors. In addition, patients with high pain behaviors in the AP group showed a clinically significant deterioration of physical impairment. This finding suggests that patients with low pain behaviors need to restructure maladaptive cognitions, whereas patients with high pain behaviors need to rebuild healthy behavior to achieve the reduction of physical impairment.

The large number of dropouts in the AP group raises a concern about the credibility of AP compared with the 2 active psychological approaches. Analysis of treatment satisfaction reported in the companion article (17), however, indicated that there were no significant differences between the groups. However, satisfaction variables in the AP group were not assessed from the dropouts.

The results of this study support the observation that patients with FMS are not homogeneous (20, 21, 35). The results indicate that treatment responders varied on important characteristics that appear amenable to treatment. Therefore, the results obtained raise the potential for improved outcomes by matching treatments to patient characteristics as originally suggested by Turk and Flor (36). One explanation for the inconsistency in results of CBT and OBT observed in the literature is that some patients were receiving treatments that were incompatible with important features present prior to treatment. The mismatch might have led to poorer outcome. Some studies have reported on the potential benefits of treatment matching (15, 36, 37), but not all studies have done so (35). Prospective studies are needed to provide a better understanding of the potential of treatment matching for patients with FMS.


  1. Top of page
  2. Abstract

Dr. Thieme 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. Thieme, Flor.

Acquisition of data. Thieme.

Analysis and interpretation of data. Thieme, Turk, Flor.

Manuscript preparation. Thieme, Turk, Flor.

Statistical analysis. Thieme, Turk.


  1. Top of page
  2. Abstract
  • 1
    Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C, Goldenberg DL, et al. The American College of Rheumatology 1990 criteria for the classification of fibromyalgia: report of the Multicenter Criteria Committee. Arthritis Rheum 1990; 33: 16072.
  • 2
    Gracely RH, Geisser ME, Giesecke T, Grant MA, Petzke F, Williams DA, et al. Pain catastrophizing and neural responses to pain among persons with fibromyalgia. Brain 2004; 127: 83543.
  • 3
    Nielson WR, Jensen MP. Relationship between changes in coping and treatment outcome in patients with fibromyalgia syndrome. Pain 2004; 109: 23341.
  • 4
    Labus JS, Keefe FJ, Jensen MP. Self-reports of pain intensity and direct observations of pain behavior: when are they correlated? Pain 2003; 102: 10924.
  • 5
    Thieme K, Spies C, Sinha P, Turk DC, Flor H. Predictors of pain behaviors in fibromyalgia patients. Arthritis Rheum 2005; 53: 34350.
  • 6
    Flor H, Knost B, Birbaumer N. The role of operant conditioning in chronic pain: an experimental investigation. Pain 2002; 95: 1118.
  • 7
    Lousberg R, Vuurman E, Lamers T, Van Breukelen G, Jongen E, Rijnen H, et al. Pain report and pain-related evoked potentials operantly conditioned. Clin J Pain 2005; 21: 26271.
  • 8
    Burckhardt DC, Goldenberg D, Crofford L, Gerwin R, Gowans S, Jackson K, et al. Guideline for the management of fibromyalgia syndrome pain in adults and children. In: APS Clinical Practice Guidelines Series, No 4. Glenview (IL): American Pain Society; 2005.
  • 9
    Nielson WR, Walker C, McCain GA. Cognitive behavioral treatment of fibromyalgia syndrome: preliminary findings. J Rheumatol 1992; 19: 98103.
  • 10
    Turk DC, Okifuji A, Sinclair JD, Starz TW. Interdisciplinary treatment for fibromyalgia syndrome: clinical and statistical significance. Arthritis Care Res 1998; 11: 18695.
  • 11
    Jacobson NS, Follette WC, Revenstorf D. Psychotherapy outcome research: methods for reporting variability and evaluating clinical significance. Behav Ther 1984; 15: 33652.
  • 12
    Kendall SA, Bartels EM, Christensent R. A meta-analysis of cognitive-behavioral therapy in fibromyalgia. J Musculoskelet Pain 2004; 12( Suppl): 4956.
  • 13
    Fordyce WE. Behavioral methods in chronic pain and illness. St. Louis: Mosby; 1976.
  • 14
    Nicassio PM, Radojevic V, Weisman MH, Schuman C, Kim J, Schoenfeld-Smith K, et al. A comparison of behavioral and educational interventions for fibromyalgia. J Rheumatol 1997; 24: 20007.
  • 15
    Thieme K, Gromnica-Ihle E, Flor H. Operant behavioral treatment of fibromyalgia: a controlled study. Arthritis Rheum 2003; 49: 31420.
  • 16
    Vlaeyen JW, Haazen IW, Schuerman JA, Kole-Snijders AM, van Eek H. Behavioural rehabilitation of chronic low back pain: comparison of an operant treatment, an operant-cognitive treatment and an operant-respondent treatment. Br J Clin Psychol 1995; 34: 95118.
  • 17
    Thieme K, Flor H, Turk DC. Psychological pain treatment in fibromyalgia syndrome: efficacy of operant behavioural and cognitive behavioural treatments (ISRCTN83953414) Arthritis Res Ther 2006; 8: R121. E-pub ahead of print. URL:
  • 18
    Al-Allaf AW, Ottewell L, Pullar T. The prevalence and significance of positive antinuclear antibodies in patients with fibromyalgia syndrome: 2–4 years' follow-up. Clin Rheumatol 2002; 21: 4727.
  • 19
    Salemi S, Rethage J, Wollina U, Michel BA, Gay RE, Gay S, et al. Detection of interleukin 1β (IL-1β), IL-6, and tumor necrosis factor-α in skin of patients with fibromyalgia. J Rheumatol 2003; 30: 14650.
  • 20
    Thieme K, Turk DC. Heterogeneity of psychophysiological stress responses in fibromyalgia syndrome patients. Arthritis Res Ther 2005; 8: R9. E-pub ahead of print. URL:
  • 21
    Turk DC, Okifuji A, Sinclair JD, Starz TW. Pain, disability, and physical functioning of patients with fibromyalgia. J Rheumatol 1996; 23: 125562.
  • 22
    Thieme K, Turk DC, Flor H. Comorbid depression and anxiety in fibromyalgia syndrome: relationship to somatic and psychosocial variables. Psychosom Med 2004; 66: 83744.
  • 23
    Flor H, Birbaumer N. Comparison of the efficacy of electromyographic biofeedback, cognitive-behavioral therapy, and conservative medical interventions in the treatment of chronic musculoskeletal pain. J Consult Clinical Psychol 1993; 61: 6538.
  • 24
    Sinclair VG, Wallstone KA. Predictors of improvement in a cognitive-behavioral intervention for women with rheumatoid arthritis. Ann Behav Med 2001; 23: 2917.
  • 25
    Okifuji A, Turk DC, Sinclair JD, Starz TW, Marcus DA. A standardized manual tender point survey. I. Development and determination of a threshold point for the identification of positive tender points in fibromyalgia syndrome. J Rheumatol 1997; 24: 37783.
  • 26
    Burckhardt CS, Clark SR, Bennett RM. The fibromyalgia impact questionnaire: development and validation. J Rheumatol 1991; 18: 72833.
  • 27
    Offenbaecher M, Waltz M, Schoeps P. Validation of a German version of the Fibromyalgia Impact Questionnaire (FIQ-G). J Rheumatol 2000; 27: 19848.
  • 28
    Kerns RD, Turk DC, Rudy TE. The West Haven-Yale Multidimensional Pain Inventory (WHYMPI). Pain 1985; 23: 34556.
  • 29
    Flor H, Rudy TE, Birbaumer N, Streit B, Schugens MM. The applicability of the West Haven-Yale Multidimensional Pain Inventory in German-speaking countries: data on the reliability and validity of the MPI-D. Der Schmerz 1990; 4: 827. In German.
  • 30
    Flor H, Behle DJ, Birbaumer N. Assessment of pain-related cognitions in chronic pain patients. Behav Res Ther 1993; 31: 6373.
  • 31
    Flor H, Heimerdinger K. Assessment of pain behavior. In: GeissnerE, JungnitschG, editors. Psychology of pain. Weinheim: Psychologie Verlags Union; 1992. p. 99106. In German.
  • 32
    Flor H, Birbaumer N. Psychobiology and interdisciplinary treatment of chronic back pain. München: GSF Forschungszentrum; 1994. In German.
  • 33
    Turk DC, Meichenbaum D, Genest M. Pain and behavioral medicine: a cognitive-behavioral perspective. New York: Guilford; 1983.
  • 34
    Williams DA. Psychological and behavioural therapies in fibromyalgia and related syndromes [review]. Best Pract Res Clin Rheumatol 2003; 17: 64965.
  • 35
    Walen HR, Cronan TA, Serbver ER, Groessl E, Oliver K. Subgroups of fibromyalgia patients: evidence for heterogeneity and an examination of differential effects following a community-based intervention. J Musculoskelet Pain 2002; 10: 932.
  • 36
    Turk DC, Flor H. Primary fibromyalgia is greater than the number of tender points: toward a multiaxial taxonomy [review]. J Rheumatol Suppl 1989; 19: 806.
  • 37
    Turk DC, Okifuji A, Sinclair JD, Starz TW. Differential responses by psychosocial subgroups of fibromyalgia syndrome patients to an interdisciplinary treatment. Arthritis Care Res 1998; 11: 397404.