Positron emission tomography in patients with fibromyalgia syndrome and healthy controls




Abnormal brain findings have previously been described in fibromyalgia syndrome (FMS) by single-photon–emission computed tomography. Our goal was to investigate change in regional cerebral glucose metabolism in people with FMS by positron emission tomography (PET) using 18F-fluorodeoxyglucose (FDG).


Twelve patients with FMS and no comorbid psychiatric diagnosis and 7 healthy pain-free controls were studied with FDG-PET in a blinded manner. Those with a psychiatric diagnosis were excluded. Brain scans were obtained using a PET scanner. Semiquantitative analysis of regional 18F-FDG uptake was performed in both cortical and subcortical brain structures.


In the resting state, there were no significant differences in 18F-FDG uptake between patients and controls for all brain structures measured.


FDG-PET scan findings in FMS were not significantly different from healthy controls. Normal results in our study may be explained by discordance between regional cerebral blood flow and regional cerebral glucose metabolism.


In recent years, functional imaging of the brain has been made possible by such techniques as positron emission tomography (PET), single-photon–emission computed tomography (SPECT), functional magnetic resonance imaging (fMRI), and magnetoencephalography (1). Most of the investigations of brain imaging on people with fibromyalgia syndrome (FMS) or FMS-related syndromes, e.g., chronic fatigue syndrome (CFS) and depression, have utilized regional cerebral blood flow (rCBF) PET and rCBF SPECT. Newer technology, e.g., fMRI, is being used more frequently. Both SPECT and PET record gamma emissions from injected radionuclides and utilize a computerized reconstruction technique that produces tomographic images of spatially distributed radionuclides in the brain (1). Cerebral blood flow is associated with brain function as a result of neuronal synaptic activity. In addition, rCBF and regional cerebral metabolism are usually coupled in the normal brain (2). Cerebral metabolism can be measured by fluorodeoxyglucose (FDG) PET. Regional cerebral blood flow is commonly measured by 15O-labeled water in PET, and by 99mTc hexamethylpropylene amine oxime in brain SPECT (3).

It has been proposed that FMS is caused by neurochemical abnormalities in the central nervous system, contributed by both aberrant physiology and psychology (4–9). FMS and related overlapping syndromes (e.g., CFS, irritable bowel syndrome, tension-type headaches, and migraine) cannot be explained by the traditional model of macroscopic and microscopic pathology (e.g., inflammation, neoplasm, or degeneration) or by a pure psychiatric model, but by a paradigm of central sensitivity-neuroendocrine dysregulation (4, 7) that embodies central sensitization (6–8), as well as perturbation of the hypothalamic–pituitary–adrenal axis (9). Functional imaging of the brain has been most useful in measuring these neurochemical abnormalities that characterize FMS and related syndromes, which we have collectively called central sensitivity syndromes (7, 8).

Among chronic pain patients, brain abnormalities, as compared with normal controls, have been described by several imaging techniques (e.g., SPECT, PET, and fMRI) in several diseases included in the central sensitivity syndromes family, e.g., CFS (10, 11), migraine (12), and irritable bowel syndrome (13). For example, patients with CFS showed multiple defects in the cerebral cortex by SPECT (10) and hypometabolism in the medial-frontal cortex and brain stem by FDG-PET (11). However, it may be noted that despite clinical similarities and a common neurohormonal pathophysiology, i.e., central sensitization (6–8), the individual pathophysiologic mechanisms in these conditions are likely to be different. Thus, the results of brain imaging in these syndromes may not be the same.

In a SPECT imaging study of 10 female patients with FMS and 7 female healthy controls, Mountz et al showed decreased rCBF in the thalamus and the caudate nucleus that was associated with low pain threshold (14). Recent controlled imaging studies in FMS have shown significantly decreased rCBF in the right thalamus and pontine tegumentum by SPECT (15); greater activation following pressure stimulus in the contralateral primary and secondary sensory cortices with ipsilateral activation of secondary sensory cortices by fMRI (16); and increased rCBF in the retrosplenial cortex and decreased blood flow in the left temporal, frontal, parietal, and occipital cortices by PET using 15O-butanol (17).

In this study, we hypothesized that FDG-PET would provide measures of regional cerebral glucose utilization in the process of central sensitization in FMS and provide complementary information to rCBF changes, expanding our understanding of cerebral activity. We undertook a blinded controlled study of brain imaging in FMS by FDG-PET under resting state. Such a PET study in FMS has not been reported previously.


Twelve white women with FMS, consecutively and newly seen in our ambulatory clinic who had consented to participate in the study and did not have a psychiatric diagnosis (see below), were included in the PET imaging study. All patients fulfilled the American College of Rheumatology criteria for FMS (18); 2 patients had also satisfied the Centers for Disease Control and Prevention criteria for CFS (19). Three patients reported one or both of the 2 neuropsychologic symptoms that are characteristic of chronic fatigue syndrome, i.e., impaired memory and impaired concentration. Patients having a concomitant disease, e.g., rheumatoid arthritis or significant osteoarthritis, were excluded. All our patients had an insidious onset of pain without a history of trauma. Complete blood count, Westergren erythrocyte sedimentation rate, serum albumin, glucose, blood urea nitrogen, creatinine, and serum glutamic pyruvic transaminase were all within normal parameters, as were serum T4 and thyroid-stimulating hormone. Seven healthy white women, all without significant pain and without a psychiatric diagnosis, were used as controls. They were evaluated in the same clinic and had similar demographic features. Five of the control subjects were either friends or neighbors of the patients, the remaining 2 were employees of our medical school.

All 12 patients had discontinued their centrally acting drugs (e.g., antidepressants and cyclobenzaprine) for 2 weeks. Acetaminophen and local heat and stretching were allowed as necessary, except for 48 hours prior to the FDG-PET scan.

All patients and controls were evaluated by a protocol that included clinical features of FMS, e.g., pain, overall fatigue, morning fatigue, sleep difficulties, and tender points among the 18 sites of the American College of Rheumatology criteria. All symptoms and tender points were graded on a 1–4 ordinal scale (1 = none, 2 = mild, 3 = moderate, 4 = severe).

To avoid probable confounding effects of psychiatric illnesses on PET scan findings, all study subjects were first screened for anxiety and depression by the State and Trait Anxiety Inventory (STAI) (20) and Zung Self-Rating Depression Scale (ZSDS) (21). Those with a STAI T-score >70 and a ZSDS index score >65 were excluded. Initially 18 consecutive patients with FMS were screened by STAI and ZSDS, but 1 of these patients had a STAI T-score >70 and 2 others had a ZSDS index scores >65. None of the 7 control subjects had these high scores. Thus, 22 subjects (15 patients and 7 controls) with STAI T-scores ≤70 and ZSDS index scores ≤65 were further evaluated for psychiatric diagnoses by our collaborating psychiatrist (SAS) by clinical psychiatric interview using Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria (22) in a blinded manner without any knowledge of the subject category. The cutoff T-score >70 for STAI and the index score >65 for ZSDS include individuals with clinically significant anxiety (14, 20) and depression (21). Only 1 patient had a diagnosis of major depression by DSM-IV criteria and was excluded from the PET study. Of the remaining 14 patients, 1 could not complete the PET procedure because of claustrophobia and another patient was excluded because she could not go without her antidepressant medication for 2 weeks. Thus, 12 patients with FMS and 7 healthy controls completed the PET imaging study.

Because physical deconditioning could possibly influence FDG-PET imaging findings, we evaluated physical activities by a validated questionnaire (23). This questionnaire includes 9 common activities (jogging, running, bicycling, lap swimming, tennis, squash, racquetball, calisthenics, or rowing). Time spent in a particular physical activity was multiplied by its typical energy expenditure expressed in metabolic equivalents (METs), providing a MET-hour score. Energy used in sitting quietly is 1 MET, which is equivalent to 3.5 ml of oxygen uptake per 1 kg of body weight per minute for a 70-kg adult. As examples, an hour of moderate-paced walking contributes 3.2 MET-hours and an hour of tennis 6 MET-hours.

FDG-PET imaging of the brain was carried out at the Downstate Clinical PET Center of the Methodist Medical Center of Central Illinois in Peoria using the Siemens/CTI (ECAT 951-031, Knoxville, TN) PET scanner with a 5–6-mm full-width-half-maximum (FWHM) in-plane resolution and a Z-axis resolution of 10.8-mm FWHM.

Study subjects fasted overnight and arrived at the PET center around 8:30 in the morning. Serum glucose was measured to exclude hyperglycemia (>125 mg/dl) prior to the intravenous injection of FDG. The patients were placed in a comfortable, relaxed position in the preparation room and 10 mCi (370 MBq) of the radiopharmaceutical 18F-FDG was injected by intravenous route. The FDG was produced by an RDS 112 cyclotron (Siemens/CTI). The tracer incorporation room was darkened and steps were taken to ensure resting and quiescent state to avoid stimulation of the auditory and visual cortices. Subjects were asked to maintain a state of neutral thoughts (14). Thirty minutes were allowed for uptake of the tracer. All subjects were then placed in a comfortable position in the PET scanner and had transmission scans performed (10 minutes) using a ring source of 68Ge to measure and correct for attenuation. Subjects then underwent a 10-minute emission scan (30–40 million counts) in 1 bed position covering 10.8 cm in axial length commencing at the vertex of the brain and extending to the base of the skull.

Images were corrected for random emissions, detector efficiency, photon attenuation, dead time, and radioactive decay. The images were reconstructed by filtered back projection with a Hann filter using a cutoff at 0.4 Nyquist. Images were sectioned in 3 planes (coronal, sagittal, and axial). Radioactivity concentrations were expressed in μCi/ml. Data were analyzed using commercial software (CTI, Knoxville, TN) on a free-standing work station and stored on optical disc.

The FDG-PET scans were normalized for differences in body surface area, and count ratios of regions of interest (ROIs) to total brain activity were computed. ROIs for count analysis were drawn to circumscribe visualized relevant structures by reference to the PET scan image. All structures of interest were visually identified by a radiologist experienced with interpretation of brain PET, and the regions were duplicated from controls to patients for consistency of ROI size and positioning. Scan analyses were performed in a blinded manner without any knowledge of the subject category, i.e., FMS versus pain-free controls, on a semiquantitative basis using mirror image of the areas of interest in defining count density in different regions in each cerebral hemisphere, including cortex, thalamus, anterior cingulate gyrus, and the basal ganglia. Regional cortical metabolism was assessed at 3 levels (vertex, base, and midcerebral). At each level, frontal, temporal, and parietal 18F-FDG uptake was calculated. Additional ROIs included the thalamus (10, 24), caudate nucleus (25), anterior cingulate gyrus, and other cortical regions implicated in the affective and cognitive aspects of pain (26).

Two-tailed t-tests and Mann-Whitney U tests were used for analyzing psychological scores of STAI and ZSDS as well as FDG-PET scan data to determine group differences. Symptoms in ordinal variables were dichotomized as moderate/severe = yes, none/mild = no, and subjected to Fisher's exact test. The results of the t-test and Mann-Whitney U test were similar. We chose to report P values by Mann-Whitney U test. Significance level was accepted a priori as ≤ 0.05.


Demographic features, important clinical features, and STAI and ZSDS scores are shown in Table 1. The MET-hours score (mean ± SD) based on physical activities among the patients and the controls in the previous month were 20.6 ± 23.3 and 19.3 ± 15.9, respectively (P = 0.44). Thus, the state of physical conditioning between the 2 groups was similar. STAI and ZSDS scores were significantly higher in the FMS group.

Table 1. Demographic, clinical, and psychologic features among the study subjects*
 FMS (n = 12)Control (n = 7)P
  • *

    FMS = fibromyalgia syndrome; NS = not significant; NA = not applicable; STAI = State and Trait Anxiety Inventory; ZSDS = Zung Self-Rating Depression Scale.

Age, mean ± SD years42 ± 742 ± 8NS
White, %100100NS
Female, %100100NS
Symptom duration, mean ± SD years9.7 ± 1.4NANA
Moderate or severe pain, %1000NA
Moderate or severe fatigue, %830< 0.001
Morning fatigue, %7500.003
No. tender points, mean ± SD15.4 ± 1.73.4 ± 1.6< 0.001
STAI-1 score, mean ± SD56.2 ± 7.442.4 ± 6.4< 0.002
STAI-2 score, mean ± SD55.3 ± 7.641.3 ± 6.3< 0.002
ZSDS score, mean ± SD51.9 ± 6.932.9 ± 3.8< 0.001

Results of the glucose metabolic rate in the ROIs by FDG-PET scan are shown in Table 2. There were no significant differences between the FMS and control groups in any region of the brain in any of the measures. The ratios of counts to total brain activity in each ROI were also quite similar between patients and controls in both sides of the brain without significant differences between the groups (data not shown). Similarly, nonsignificant results were obtained when the count differences between the right and the left hemispheres were compared between patients and controls (data not shown). A separate analysis after elimination of 2 patients with FMS as well as CFS showed almost identical results.

Table 2. FDG-PET imaging findings*
 FMS (n = 12)Control (n = 7)P
  • *

    Data presented as mean ± SD count in μCi/ml. FMS = fibromyalgia syndrome.

Right caudate nucleus   
 Total count9.24 ± 1.2510.40 ± 2.740.33
 Average count0.88 ± 0.120.99 ± 0.260.33
 Surface area57.84 ± 0.0057.84 ± 0.001.00
 Volume195.20 ± 0.00195.20 ± 0.001.00
Left caudate nucleus   
 Total count9.03 ± 1.4110.06 ± 2.250.23
 Average count0.86 ± 0.130.96 ± 0.210.23
 Surface area57.84 ± 0.0057.84 ± 0.001.00
 Volume195.20 ± 0.00195.20 ± 0.001.00
Right thalamus   
 Total count24.37 ± 4.9824.69 ± 5.530.98
 Average count1.03 ± 0.211.05 ± 0.230.90
 Surface area129.45 ± 0.00129.45 ± 0.001.00
 Volume436.88 ± 0.00436.88 ± 0.001.00
Left thalamus   
 Total count23.84 ± 4.7624.76 ± 6.150.72
 Average count1.01 ± 0.201.05 ± 0.260.72
 Surface area129.45 ± 0.00129.45 ± 0.001.00
 Volume436.88 ± 0.00436.88 ± 0.001.00
Right basal ganglia   
 Total count61.78 ± 12.3467.22 ± 16.990.43
 Average count0.88 ± 0.180.92 ± 0.240.70
 Surface area384.33 ± 0.00403.10 ± 0.000.36
 Volume1,297.09 ± 0.001,360.44 ± 0.000.36
Left basal ganglia   
 Total count61.71 ± 12.0366.53 ± 18.810.50
 Average count0.88 ± 0.170.90 ± 0.250.80
 Surface area384.33 ± 0.00403.10 ± 0.000.36
 Volume1,297.09 ± 0.001,360.44 ± 0.000.36
Right insula   
 Total count14.97 ± 3.3914.61 ± 3.700.83
 Average count0.99 ± 0.220.97 ± 0.240.83
 Surface area82.63 ± 0.0082.63 ± 0.001.00
 Volume387.84 ± 0.00387.84 ± 0.001.00
Left insula   
 Total count14.17 ± 2.9314.65 ± 2.700.73
 Average count0.94 ± 0.190.97 ± 0.180.73
 Surface area60.59 ± 0.0060.59 ± 0.001.00
 Volume284.42 ± 0.00284.42 ± 0.001.00
Right anterior cingulate gyrus   
 Total count11.03 ± 2.4711.10 ± 3.290.96
 Average count1.00 ± 0.221.00 ± 0.290.96
 Surface area60.59 ± 0.0060.59 ± 0.001.00
 Volume284.42 ± 0.00284.42 ± 0.001.00
Left anterior cingulate gyrus   
 Total count10.99 ± 2.1911.40 ± 2.790.73
 Average count0.99 ± 0.191.03 ± 0.250.73
 Surface area60.59 ± 0.0060.59 ± 0.001.00
 Volume284.42 ± 0.00284.42 ± 0.001.00
Right frontal cortex   
 Total count142.16 ± 28.14139.68 ± 32.970.86
 Average count0.96 ± 0.190.94 ± 0.220.86
 Surface area816.61 ± 0.00816.61 ± 0.001.00
 Volume2,756.10 ± 0.002,756.10 ± 0.001.00
Left frontal cortex   
 Total count141.14 ± 30.14144.64 ± 27.420.80
 Average count0.95 ± 0.200.98 ± 0.180.80
 Surface area816.61 ± 0.00816.61 ± 0.001.00
 Volume2,756.10 ± 0.002,756.10 ± 0.001.00
Right temporal cortex   
 Total count73.84 ± 12.4571.68 ± 13.330.73
 Average count1.05 ± 0.181.02 ± 0.200.73
 Surface area386.96 ± 0.00386.96 ± 0.001.00
 Volume1,306.00 ± 0.001,306.00 ± 0.001.00
Left temporal cortex   
 Total count71.13 ± 14.1672.48 ± 12.350.84
 Average count1.01 ± 0.201.03 ± 0.180.84
 Surface area386.96 ± 0.00386.96 ± 0.001.00
 Volume1,306.00 ± 0.001,306.00 ± 0.001.00
Right parietal cortex   
 Total count84.66 ± 17.9087.26 ± 16.660.76
 Average count0.91 ± 0.190.94 ± 0.180.76
 Surface area513.65 ± 0.00513.65 ± 0.001.00
 Volume1,733.60 ± 0.001,733.60 ± 0.001.00
Left parietal cortex   
 Total count81.93 ± 16.0987.65 ± 15.110.46
 Average count1.53 ± 2.250.94 ± 0.160.39
 Surface area513.65 ± 0.00513.65 ± 0.001.00
 Volume1,733.60 ± 0.001,733.60 ± 0.000.46


Ours is the first controlled and blinded study of FDG-PET imaging in FMS, to our knowledge. We believe our study design was satisfactory. There were no significant differences between patients and healthy, pain-free controls having the same or similar sex, race, age, psychological status, and physical activities.

FDG-PET scan findings in FMS were not significantly different from those among the healthy controls. However, it should be noted that our study had certain limitations that might not have provided maximum sensitivity to detect subtle changes in 18F-FDG metabolism. Unlike the SPECT scan methods employed to investigate FMS by Mountz et al (14), images of brain PET scan with coalignment to the individual patient's magnetic resonance images of brain structures was not performed. Localization of a brain structure in our study depended on visual interpretation of the FDG-PET. However, our method of assessing 18F-FDG changes between patients and controls was carefully coregistered, and statistically significant differences, if present, would have been revealed.

It must be considered that imaging rCBF and metabolism may not produce parallel changes within the same brain structure. In another study of 18 patients with CFS comparing rCBF with metabolism, only 3 patients showed abnormal results by FDG-PET, as compared with 15 patients by SPECT (27). The authors conclude that there is a discordance between SPECT brain perfusion and FDG-PET uptake, and it is possible to have brain perfusion abnormalities without corresponding changes in glucose uptake (27).

FDG-PET studies have been found to be abnormal in other neurochemically based diseases, such as major depression (28, 29), generalized anxiety disorder (30), and obsessive-compulsive disorder (31). In some of these disorders, discordance between rCBF and glucose metabolism has been shown (27). In depression, decreased uptake of FDG-PET in the frontal and parietal cortex has been reported (29), whereas FDG-PET studies in obsessive-compulsive disorder are characterized by increased uptake of 18F-FDG in the orbital frontal gyrus and caudate nucleus (31). Our normal FDG-PET results in FMS would suggest that fibromyalgia is different from psychiatric diseases, as we had previously suggested (4).

Functional neuroimaging of the brain is a new important tool to study the biophysiology of chronic pain conditions, including FMS. However, all functional imaging techniques are not the same and results may not be comparable. Future imaging studies should address the issue of optimum sensitivity and specificity using various techniques, and utilizing a large sample size. Stimulation studies (16) are also important. Cerebral blood flow studies by 15O-PET are dynamic procedures that evaluate changes in perfusion occurring over a short period of time. In contrast, FDG-PET images a regional cerebral metabolism integrated over a longer time. In conclusion, our FDG-PET study in FMS showed no significant difference in 18F-FDG uptake between patients and controls for all brain structures measured. Other brain imaging techniques may assist to elucidate the neurophysiology of FMS.