Content uploaded by Dane B Cook
Author content
All content in this area was uploaded by Dane B Cook on Nov 10, 2017
Content may be subject to copyright.
The Journal of Rheumatology 2004; 31:2
364
From the Chronic Fatigue Syndrome Cooperative Research Center and
Departments of Radiology, Psychiatry, and Neurosciences, University of
Medicine and Dentistry of New Jersey, New Jersey Medical School,
Newark; and the War-Related Illnesses and Injury Study Center, Veterans
Affairs New Jersey Health Care, East Orange, New Jersey, USA.
Supported by the pilot grants module of the National Institutes of Health,
grant AI-32247.
D.B. Cook, PhD, Assistant Professor, Department of Radiology, CFS
Cooperative Research Center, War-Related Illness and Injury Study
Center; G. Lange, PhD, Associate Professor, Department of Radiology,
War -Related Illness and Injury Study Center; D.S. Ciccone, PhD,
Assistant Professor, Department of Psychiatry; W-C. Liu, PhD, Assistant
Professor, Department of Radiology; J. Steffener, MA, Department of
Psychiatry; B.H. Natelson, MD, Professor, Department of Neurosciences,
Director, CFS Cooperative Research Center, Director, War-Related Illness
and Injury Study Center.
Address reprint requests to Dr. D.B. Cook, Fatigue Research Center, New
Jersey Medical School, 88 Ross Street, East Orange, NJ 07018. E-mail:
cookdb@njneuromed.org
Submitted August 30, 2002; revision accepted June 26, 2003.
Fibromyalgia (FM) is a condition characterized by wide-
spread musculoskeletal pain and multiple tender point sites1.
The etiology of the syndrome is unknown, and no consistent
underlying mechanism has been identified. However,
evidence of increased pain sensitivity and altered nocicep-
tive processing in patients with FM suggests that dysregula-
tion of the central nervous system (CNS) may be an impor-
tant element underlying FM pain.
Results from studies examining sensitivity to experimen-
tally induced pain have shown that patients with FM have
lower pain thresholds and report higher pain ratings in
response to pressure, heat, cold and electrical stimuli2-6.
Experiments examining pain regulatory systems have
shown patients with FM to display a dysregulation of diffuse
noxious inhibitory controls7,8, an exaggerated wind-up
response to repetitive pain stimuli9, and absence of an exer-
cise-induced analgesic response5,10. Together, these results
point to central dysregulation of the nociceptive system.
However, one limitation of previous research examining
nociceptive processes in FM has been the reliance on self-
report measures of pain. Objective evidence of abnormal
nociception is needed to better understand the pathophysio-
logic processes involved, and provide converging evidence
of the patient’s self-reported symptoms.
The emergence of brain imaging as an investigative tool
has resulted in a greater understanding of the complexity of
the nociceptive system in humans. Research using experi-
mental pain stimuli (e.g., noxious heat, noxious chemicals,
electricity), and employing positron emission tomography
(PET) and functional magnetic resonance imaging (fMRI)
Functional Imaging of Pain in Patients with
Primary Fibromyalgia
DANE B. COOK, GUDRUN LANGE, DONALD S. CICCONE, WEN-CHING LIU, JASON STEFFENER,
and BENJAMIN H. NATELSON
ABSTRACT. Objective. To examine the function of the nociceptive system in patients with fibromyalgia (FM)
using functional magnetic resonance imaging (fMRI).
Methods. Two groups of women, 9 with FM and 9 pain-free, volunteered to participate. In
Experiment 1, we assessed psychophysical responses to painful stimuli and prepared participants for
fMRI testing. For Experiment 2, subjects underwent fMRI scanning while receiving painful and
nonpainful heat stimuli. Conventional and functional MR images were acquired using a 1.5 T MR
scanner. Scanning occurred over 5 conditions. Condition 1 served as a practice session (no stimuli).
Conditions 2 and 5 consisted of nonpainful warm stimuli. Conditions 3 and 4 consisted of an
absolute thermal pain stimulus (47°C) and a perceptually equivalent pain stimulus delivered in coun-
terbalanced order.
Results. Experiment 1 indicated that subjects with FM were significantly more sensitive to experi-
mental heat pain than controls (p < 0.001). In Experiment 2, fMRI data indicated that the FM group
exhibited greater activity than controls over multiple brain regions in response to both nonpainful
and painful stimuli (p < 0.01). Specifically, in response to nonpainful warm stimuli, FM subjects had
significantly greater activity than controls in prefrontal, supplemental motor, insular, and anterior
cingulate cortices (p < 0.01). In response to painful stimuli, FM subjects had greater activity in the
contralateral insular cortex (p < 0.01). Data from the practice session indicated brain activity in pain-
relevant areas for the FM group but not for controls.
Conclusion. Our results provide further evidence for a physiological explanation for FM pain.
(J Rheumatol 2004;31:364–78)
Key Indexing Terms:
FIBROMYALGIA CENTRAL NERVOUS SYSTEM
MUSCULOSKELETAL PAIN BRAIN IMAGING
techniques, has identified many of the brain areas involved
in processing the nociceptive signal in healthy people. The
areas most consistently implicated in pain processing, and
thus considered pain-relevant, are the sensory cortex,
prefrontal cortex, inferior parietal cortex, anterior cingulate
cortex, insula, lentiform nucleus, and thalamus11-16.
Moreover, improvements in behavioral research designs that
control attention during scanning, examine painful versus
nonpainful stimuli, and obtain pain intensity and affective
ratings during stimulation have provided important informa-
tion regarding the cognitive, sensory, and emotional
processes that are inherently involved in the perception of
pain12,15,17-21.
Recently, 2 single-photon emission computed tomog-
raphy (SPECT) studies in FM have reported reduced
regional cerebral blood flow (rCBF) to the thalamus, heads
of the caudate nucleus, and pontine tegmentum at rest22,23.
These results are consistent with data obtained from other
clinical pain populations (e.g., cancer pain, neuropathic
pain, reflex sympathetic dystrophy, rheumatoid arthritis)
and have been interpreted as an inability of the nociceptive
system to modulate or compensate for the constant barrage
of incoming nociceptive signals24-27. The results of these
studies have provided important information regarding the
resting state of the brain in chronic pain. To our knowledge,
only one report exists using functional neuroimaging
methods to assess how the brains of patients with FM
respond to painful stimuli. In a well designed study, Gracely,
et al28 reported that fMRI brain responses to experimental
pressure pain, set at either similar stimulus levels or similar
subjective pain levels, were augmented in FM patients
compared to controls. This finding supports the view that
physiological processing of pain is altered in FM.
We used fMRI to examine neural activation patterns to
experimental pain in patients with FM and healthy controls.
We attempted to exploit recent improvements in fMRI
research by controlling attention during scanning, using
both painful and nonpainful stimuli, obtaining pain ratings
during stimulation rather than after stimulus presentation,
and controlling for perceptual differences in pain (stimulus)
intensity. Our design as detailed in Table 1 was intended to
test the following hypotheses: (1) presented with a
nonpainful (warm) stimulus, FM patients will exhibit neural
responses in brain regions associated with pain perception
whereas controls will not; (2) presented with a painful (hot)
stimulus of equal absolute magnitude, FM patients will
exhibit greater neural responses in brain regions associated
with pain perception compared to controls; (3) presented
with a perceptually equivalent pain (hot) stimulus, FM
patients and controls will display similar neural responses in
brain regions associated with pain perception; and (4) neural
responses to nonpainful (warm) stimuli will remain elevated
in FM patients but not in controls following the presentation
of several painful stimuli (i.e., fail to recover). If supported,
the results would provide neurophysiological evidence of
altered sensory processing and provide objective support for
enhanced pain perception in patients with FM.
MATERIALS AND METHODS
This study consisted of 2 separate experiments conducted about one week
apart. Experiment 1 was intended to assess psychophysical responses to
painful stimuli and to prepare the participants for subsequent fMRI testing.
This included practice at rating painful and nonpainful heat stimuli using
the same equipment used in the fMRI experiment, and establishing temper-
atures to be used during fMRI testing. For Experiment 2, participants
reported to the MRI suite of the university hospital for fMRI data collec-
tion.
Subjects. Eighteen right handed females (n = 9 FM, n = 9 healthy controls)
between the ages of 18 and 45 years volunteered to participate. All partici-
pants were recruited from a large patient pool through the CFS/FM Center.
Patients met American College of Rheumatology criteria for FM1and had
pain as their major symptom complaint. FM patients also met US Centers
for Disease Control criteria for chronic fatigue syndrome29. Participants
were screened to ensure that they were not claustrophobic or pregnant, and
had no metal in the body. Participants were also screened to ensure that they
were not taking anticonvulsant, antihypertensive, antidepressant, or pain
medication for at least 3 weeks prior to the study. Additionally, participants
were assessed using the Computerized Diagnostic Interview Schedule (Q-
DIS)30 to confirm they were free from Axis I psychiatric disorders that
might affect pain ratings. Controls were healthy females, free from any
current or lifetime Axis I psychiatric diagnosis, and not taking medication
other than oral contraceptives. All participants signed an informed consent
explaining the procedures of the study prior to testing.
Experiment 1
Procedures. Participants completed a battery of questionnaires including
the Spielberger State-Trait Anxiety Inventory (STAI), Beck Depression
Inventory (BDI), Kohn Reactivity Scale (KRS), and the short-form McGill
Pain Questionnaire (MPQ)31-34. The STAI and BDI were used to assess
anxiety and depression prior to testing, and to examine possible relation-
ships between affect and pain ratings. We employed the KRS because of
reports of hypervigilance in patients with FM that may affect responses to
sensory testing35. The MPQ was employed to assess current pain and to
examine the effect of chronic pain symptoms on acute pain ratings.
Sensory stimulation apparatus. Painful and nonpainful heat stimuli for all
tests were delivered using a quantitative sensory testing unit (Medoc TSA-
2001, Medoc Ltd., Ramat Yishai, Israel). Stimuli were delivered to the
thenar eminence of the nondominant hand with a 300 mm2Peltier ther-
mode, and using the WinTSA software provided by Medoc.
Threshold and suprathreshold testing. Heat and pain thresholds were
assessed using the ascending method of limits. Baseline temperature for the
thermode was kept at 32°C. The rate of temperature change was set to
Cook, et al: Functional imaging of FM pain 365
Table 1. Order of heat and pain stimuli during fMRI scanning.
Run Condition Description
1Practice Practice rating on the pain intensity scale;
no stimulation
2Random warm 5 random warm stimuli from 34˚C to 42˚C
stimuli in 2˚ increments
3 and 4 Pain stimuli Counterbalanced presentation of either 47˚C
or pain previously rated as a 5 on the 0–10 scale
5Random warm 5 random warm stimuli from 34˚C to 42˚C
stimuli in 2˚ increments
increase at 1°C/s with a return to baseline rate of 10°C/s for both the warm
threshold and pain threshold measurements.
Psychophysical responses to pain were assessed using 10 random
suprathreshold heat stimuli. Each stimulus lasted 10 s with an interstimulus
interval of 2 min. Temperatures ranged from 44.5°C to 49°C in 0.5°C incre-
ments. Stimuli were rated along 2 separate 0–10 category-ratio (CR-10)
scales designed to measure both pain intensity and pain unpleasantness36
(Figure 1). The CR-10 scales have been widely used in the study of pain
and have been shown to be reliable and valid measures of pain ratings37-41.
Moreover, the CR-10 scale with the specific placement of the verbal
anchors was designed to combine the advantages of ratio-scaling tech-
niques with those of simple rating methods, thus providing estimations of
psychophysical growth functions while maintaining the ability to make
interindividual comparisons37. A detailed description of the methods used
to determine the psychophysical characteristics of the painful stimuli is
provided in an Appendix. Immediately after the sensory testing session, the
State Anxiety Inventory (SAI) form and MPQ were given to assess the level
of anxiety provoked by testing and to describe the heat pain experienced
during the testing session, respectively.
Experiment 2
Participants arrived at the MRI suite and completed the STAI, BDI, and
MPQ for the same purpose as described in Experiment 1. Participants were
then reacquainted with the CR-10 scales. Immediately prior to testing,
subjects were reminded that they were about to receive a number of stimuli,
some of which may be considered “extremely painful,” but that no skin
damage would result.
Imaging parameters. Conventional and fMRI image acquisitions were
performed on a 1.5 T Echospeed Horizon MR scanner (GE, Milwaukee,
WI, USA) using whole-body echoplanar imaging (EPI) with a whole-head
transmit-receive coil. Foam cushions were used to immobilize the head
within the coil to minimize motion degradation. The subjects wore MRI-
compatible earmuffs for communication with the experimenter and to mini-
mize scanner noise during acquisition. Subjects were then fitted with a set
of fiber optic goggles (Avotec, Jensen Beach, FL, USA) for viewing the
CR-10 scales (Appendix). Conventional acquisitions consisted of twenty-
eight 5 mm thick T1 weighted [TR 450 ms, TE 20 ms, field of view (FOV)
24 cm] axial images with a matrix of 256 ×256. Functional MRI acquisi-
tions were obtained with a gradient echo sequence and a spiral trajectory
(TR 2000, TE 40) and consisted of twenty 5 mm thick slices. The acquisi-
tion matrix was 64 ×64 and FOV 20 cm2, delivering an in-plane voxel reso-
lution of 3.125 ×3.125 ×5 mm. The images were acquired at the same slice
locations for subsequent overlays to correlate anatomic location with
changes in fMRI signal intensity. The total experimental session, including
fMRI and conventional MRI acquisitions, was roughly 40 min.
Experimental protocol. After T1 image acquisition, subjects were fitted
with the thermode on the thenar eminence of the left hand. A toggle switch
used to rate pain intensity and unpleasantness was placed in the right hand.
The toggle switch consisted of a small dial (~10 mm diameter) that was
easily moved with use of the subject’s thumb and index finger. The amount
of movement required to scroll the entire scale was less than 1 inch.
Subjects were instructed not to move during the positioning of these
devices.
Functional images were acquired over 5 separate runs using a boxcar
design (Figure 2). Each run consisted of six 30 s off-periods (no stimula-
tion) and five 10 s on-periods (heat stimuli) for a total of 230 s. The pain
intensity scale was in view for the subject throughout the experimental
protocol, and the pain unpleasantness scale was shown at the end of each
run. The order and description of each condition is given in Table 1 and a
depiction of the experiment design in Figure 2.
After functional MR data acquisition, the raw data were transferred to
a remote Sun SPARC10 workstation through a dedicated Ethernet connec-
tion. The raw data were reconstructed offline using software developed
under IDL (Interactive Data Language, Research Systems, Boulder, CO,
USA). Due to the saturation effect, the first 3 functional images of each run
were not used in statistical analyses. Images were realigned for motion
correction, coregistered to the T1 anatomical image, and normalized to
The Journal of Rheumatology 2004; 31:2
366
Figure 1. Pain intensity and pain unpleasantness scales. Subjects viewed the pain intensity scale throughout each condition,
while the pain unpleasantness scale was viewed immediately after each condition.
Talairach space using Statistical Parametric Mapping software (SPM99)42.
The data were high-pass filtered with a 128 s cutoff period, low-pass
filtered with an 8 s cutoff period, smoothed spatially (8 ×8 ×10 mm), and
globally scaled to normalize the signal intensity. The maximum displace-
ment was measured using the value (mm or degree) from the motion
correction in x, y, z, roll, pitch, and yaw directions. Immediately after the
fMRI testing session, the SAI and MPQ were given to assess the level of
anxiety provoked by testing and to describe the heat pain experienced
during the testing session, respectively.
Statistical analyses. For Experiment 1, subject characteristics, question-
naire data, thresholds, and peak ratings were analyzed using independent
samples t tests with Bonferroni correction for multiple comparisons. Due to
technical error or pain tolerance limitations, 4 subjects with FM did not
receive the 48.5°C and 49°C stimuli, and therefore data analyses were
performed on the 8 random stimuli that were administered to all partici-
pants. Ratings of pain intensity and unpleasantness for the 8 random heat
stimuli were examined using a 2 group (FM and Control) ×8 temperature
(44.5°C to 48°C) repeated measures analysis of variance. Analysis of
covariance (ANCOVA) were used to examine main and interaction effects
after parceling out the effects of selected potential confounding variables.
Linear regression analysis and curve estimation was performed on natural
log plots to describe the relationship between pain ratings and temperature
for each group. When appropriate, effect sizes were calculated according to
the method of Cohen43. Rough guidelines for Cohen’s d are that 0.2 is
small, 0.5 is moderate, and 0.8 is large. For Experiment 2, questionnaire
data and unpleasantness ratings were examined using independent samples
t tests. Pain intensity ratings during fMRI testing were examined using a 2
group ×5 temperature repeated measures ANOVA.
Analysis of the fMRI data was performed using SPM99 software42.
First, single-subject analyses were performed by linear regression of the
fMRI data. Within-group analyses were then conducted by examining the
mean responses during each condition for the FM and control groups sepa-
rately, using dependent samples t tests. The statistical activation maps were
viewed using an uncorrected p value of 0.01. This voxel-based analysis
searched the entire brain for areas that have been previously reported as
pain-relevant. These regions included the somatosensory, motor, prefrontal,
parietal, insular, and cingulate cortices, the thalamus, amygdala, and basal
ganglia11-16,22,23. We then compared the FM and control groups using region
of interest (ROI) analyses. These between-groups comparisons were
restricted to pain-relevant regions that reached statistical significance in the
within-group analyses. This process reduced the search volumes from >
100,000 voxels for the within-group analyses to less than 3000 voxels for
the between-groups ROI analyses. This procedure reduced the probability
of making a type II error by reducing the number of comparisons. The ROI
analyses (p < 0.01, Bonferroni corrected) tested for differences between
groups within pain-relevant areas after controlling for depressed mood
(BDI scores) and chronic pain (MPQ total scores). Finally, regression
analyses were used to determine the relationship between pain ratings
during scanning and brain activity for the entire sample.
Since the heat stimulus was delivered to the left thenar eminence,
contralateral activity represents right brain activity, while ipsilateral
activity represents left brain activity. Locations, z-scores, and cluster values
for Tables 2 and 3 represent the numbers observed for the maximal voxel
within the gray matter of a given cluster. Table and text data are presented
as mean ± SD and graphic data are presented as mean ± SE.
RESULTS
Experiment 1
Subject characteristics. There were no significant differ-
ences in the age, height, or weight of the 2 groups. FM
participants were a mean age of 37 ± 5 years, height 163 ±
6 cm, and weight 67 ± 19 kg. Controls were a mean age 35
±3 years, height 163 ± 6 cm, and weight 65 ± 11 kg. Prior
to testing, the FM participants reported more depressed
mood (BDI scores: FM 8.4 ±7 vs controls 2.4 ±2; t1,16 = 2.6;
p = 0.02), were more reactive (KRS scores: FM 87.6 ± 14 vs
controls 66.3 ± 15; t1,16 = 3.1; p = 0.007), and were in more
resting pain than controls (MPQ total: FM 10.3 ± 7 vs
controls 0.7 ± 1; t1,16 = 4.0; p = 0.001). There were no signif-
icant differences in state or trait anxiety prior to testing.
Thresholds and peaks. There were no significant group
differences for either the temperatures indicated as warm
threshold (FM 33.3 ± 0.5°C vs controls 33.4 ± 0.7°C) or
pain threshold (FM 43.5 ± 2.6°C vs controls 44.5 ± 3.4°C).
The effect size for the difference in pain threshold tempera-
ture was moderate (ES d = 0.36). Peak pain intensity at 48°C
Cook, et al: Functional imaging of FM pain 367
Figure 2. Agraphic display of the experiment design used in this study. Each session consisted of 5
runs representing different conditions of the study. Each run lasted a total of 230 s and was divided
into six 30 s “off-periods” and five 10 s “on-periods.”
was significantly greater in the FM group compared to
controls (FM 5.0 ± 2.8 vs controls 2.9 ± 1.2; p = 0.05). Peak
pain unpleasantness at 48°C was not different between
groups (FM 4.1 ± 2.8 vs controls 3.7 ± 2.3).
Suprathreshold pain testing. Pain intensity ratings for the 8
stimuli are presented in Figure 3. Repeated measures
ANOVA revealed a significant group ×temperature interac-
tion (F7,112 = 13.2, p < 0.001] for pain intensity ratings.
Linear regression of the log-transformed data revealed that
for the control group each 0.5°C increase in temperature was
associated with an 18% increase in pain intensity ratings.
For the FM group each increase in temperature was associ-
ated with a 35% increase in pain intensity ratings. Thus, the
FM group was almost twice as sensitive to painful heat
stimuli as controls. Separate ANCOVA using the total score
from the BDI and MPQ, respectively, failed to eliminate the
interaction, suggesting that depression and resting pain did
not explain the increased pain sensitivity observed in the
FM group. There was neither significant main effects nor a
significant interaction for pain unpleasantness ratings.
Post-test pain and anxiety. After pain testing, FM partici-
pants reported being significantly more anxious than
controls (STAI scores: FM 30.2 ± 3 vs controls 25.2 ± 4; t1,16
= 3.2; p = 0.006). MPQ totals describing the heat pain expe-
rienced during testing were higher in the FM group (ES d =
0.64), but were not significantly different from controls (FM
8.3 ± 5.4 vs controls 5.6 ± 3.1; p = 0.09). This was likely due
to the high variability within groups. Further analysis of
pain descriptors indicated that the 3 most common adjec-
tives used by both groups were “hot-burning,” “sharp,” and
“tender.” Thus, both FM and control participants used
similar terms to describe the pain experience.
Experiment 2
Before testing, FM participants reported a more depressed
mood (BDI scores: FM = 7.7 ± 7 vs controls 0.9 ± 1; t1,16 =
2.9; p = 0.01) and were in greater resting pain (MPQ total:
FM 7.7 ± 6 vs controls 0.7 ± 1; t1,16 = 3.5; p = 0.003)
compared to controls. There were no significant group
differences in anxiety prior to scanning.
Functional imaging results. Significant blood oxygen level-
dependent (BOLD) signal increases (p < 0.01) for each
condition can be seen in Tables 2 (control group) and 3 (FM
group). The combined results across all experimental condi-
tions are illustrated in Figure 4.
Practice session (Condition 1): within-group analyses. SPM
analysis indicated no significant areas of activation in the
control group (Table 2); however, this was not the case for
The Journal of Rheumatology 2004; 31:2
368
Table 2. Local maxima, z scores, cluster sizes, and associated gray matter regions expressed as Brodmann’s classifications or anatomical region for the control
group. Activations within the control group across conditions. Coordinates (x, y, z) represent the most significantly activated voxel based on the highest z
score. Depending on the size of the cluster, several anatomical areas may be represented. Regions within each cluster were identified with locally developed
software designed to examine each voxel within a cluster for its designated area based on Talairach coordinates.
Condition Control Group
Local Z Score Cluster Gray Matter Regions within the Clusters
Max (x, y, z) Size (KE)Left Right
Practice No significant areas No significant regions No significant regions
Random warm Stimuli #1 4 –18 12* 2.7 16 No significant regions Thalamus
Pain: 47˚C stimulus 42 –2 44** 3.6 25 BA 6
–18 12 4** 3.4 16 Putamen
32 42 34** 3.2 13 BA 9
20 16 –4** 3.0 13 Putamen
–2 –16 10** 3.0 12 Thalamus
20 2 36** 3.0 12 BA 24
54 –30 48* 2.6 6 BA 2
–30 34 28* 2.6 10 BA 9
Pain: temperature rated as 5 40 16 –2** 3.7 15 Anterior insula
62 –10 6** 3.5 141 BA 6, BA 40
32 42 34* 2.9 15 BA 9
–38 –22 10* 2.8 5 Anterior insula
–20 –24 42* 2.9 5 BA 31
–24 56 24* 2.8 3 BA 10
Random warm stimuli #2 0 –58 40** 3.9 34 BA 7 BA 7
40 34 18** 3.9 83 BA 10
4 –28 2** 3.5 156 Thalamus Thalamus
8 –26 36** 3.3 54 BA 23, BA 24, BA 31
4 –34 50** 3.2 18 BA 5
10 14 40** 3.1 24 BA 32
42 – 16 2* 2.8 11 Anterior insula
* p ≤0.005, ** p ≤0.001.
the FM group (Table 3). Practice ratings and placement of
the thermode on the hand resulted in significant activation
bilaterally in the supplemental motor (BA 6), primary motor
(BA 4), and sensory association (BA 5) cortices. Significant
contralateral (right side) activity was observed in prefrontal
(BA 9), superior parietal (BA 7), and posterior insular
cortices. Ipsilateral activity was identified in the putamen.
Between-group analyses. The control group did not exhibit
any significant activation during practice. Therefore, there
were no comparisons to the FM group. T tests of the areas
of activation for the FM group (FM – controls) indicated no
significantly greater areas when compared to controls. Thus,
while activity was observed in several pain-relevant areas in
the FM group, these areas were not significantly greater than
Cook, et al: Functional imaging of FM pain 369
Table 3. Local maxima, z scores, cluster sizes, and associated gray matter regions expressed as Brodmann’s classifications or anatomical regions for the FM
group. Activations within the control group across conditions. Coordinates (x, y, z) represent the most significantly activated voxel based on the highest z
score. Depending on the size of the cluster, several anatomical areas may be represented. Regions within each cluster were identified with locally developed
software designed to examine each voxel within a cluster for its designated area based on Talairach coordinates.
Condition FM Group
Local Z Score Cluster Gray Matter Regions within the Clusters
Max (x, y, z) Size (KE)Left Right
Practice 0 –34 76** 3.7 265 BA 4, BA 5, BA 6 BA 4, BA 5, BA 6
–54 –2 24** 3.6 30 BA 4, BA 6
–40 –8 –6** 3.3 107 Putamen
30 16 26** 3.2 40 BA 9
20 –72 50* 2.8 35 BA 7
30 –52 62* 2.7 31 BA 7
34 4 16* 2.7 11 Posterior insula
Random warm stimuli #1 – 2 28 10** 4.9 1422 BA 6, BA 8, BA 24, BA 32 BA 6, BA 8, BA 9, BA 10
Caudate BA 24, BA 32, caudate,
putamen, globus pallidus
–40 14 40** 4.5 355 BA 6, BA 9
–16 52 18** 4.2 261 BA 9, BA 10
4 –32 60** 4.0 77 BA 4, BA 5, BA 6
–44 –26 4** 3.4 406 Posterior insula
44 –20 22** 3.0 74 Posterior insula
–12 32 46* 3.0 127 BA8
30 –30 62* 3.4 11 BA 3
Pain: 47˚C stimulus 40 –14 22** 4.5 2028 BA 3, BA 4, BA 6, anterior &
posterior insula
–4 4 4** 4.0 332 Caudate Putamen, globus pallidus, caudate
56 10 32** 4.0 141 BA 9
–56 –42 34** 3.3 75 BA 40
44 –30 30** 3.2 46 BA 2, BA 40
0 6 26** 3.1 70 BA 24, BA 32 BA 24, BA32
24 –24 4** 3.1 33 Thalamus
–4 30 28** 3.1 320 BA 6, BA 9, BA 32 BA 8, BA 32
–40 –16 50* 2.8 20 BA 4
Pain: temperature rated as 5 –40 16 8** 3.7 377 BA 9, anterior insula
–54 –30 16** 3.6 35 BA 40
–44 44 10** 3.1 24 BA 10
64 –28 18* 2.8 19 BA 40
34 16 12* 2.7 10 Anterior insula
Random warm stimuli #2 –32 18 6** 4.3 646 BA 10, anterior insula
–4 8 52** 4.0 325 BA 6, BA 8, BA 24, BA 32
–10 –2 12** 3.7 188 Thalamus Thalamus
48 –58 36** 3.7 88 BA 40
22 40 36** 3.6 177 BA 8, BA 9
–14 48 32** 3.4 42 BA 9
–10 56 20** 3.3 24 BA 10
12 2 60** 3.1 31 BA 6
46 –36 18** 3.1 125 Posterior insula
–30 16 40** 3.0 93 BA 8
12 12 –2** 2.9 19 Caudate, putamen
0 –70 30** 2.9 78 BA 7 BA 7
26 0 0* 2.5 5 Putamen
* p ≤0.005, ** p ≤0.001.
the control group. Further, as shown in Figure 5, neither the
BOLD signals nor the design matrix was related to the
measured movement variables.
Do FM patients exhibit pain-relevant activation to
nonpainful warm stimuli? (Condition #2): perceptual
ratings. There were no significant differences between
groups for either the pain intensity or unpleasantness ratings
for the 5 warm stimuli. Both groups generally rated the
warm stimuli as not painful. The average pain rating for the
warm stimuli was 0.5 (± 0.2) and 0.5 (± 0.3) for the FM and
control groups, respectively. The average unpleasantness
rating for the FM group was 1.3 (± 0.5) and for the control
group 0.8 (± 0.4).
Within-group analyses. Warm stimulation resulted in signif-
icant activation in the contralateral dorsomedial nucleus of
the thalamus in controls (t = 3.7, p = 0.004). The FM group
exhibited bilateral activity in the prefrontal cortex (BA9,
BA10), supplemental motor area (BA 6, BA 8), anterior
cingulate cortex (BA 24, BA 32), the posterior insula, and
the corpus striatum encompassing the caudate, putamen and
a section of the globus pallidus. Contralateral activity was
seen in the primary motor (BA 4) and sensory (BA 3)
cortices.
Between-group analyses. The thalamic activation observed
in the controls was not significantly different from the FM
group. The FM group exhibited significantly greater activity
bilaterally (p < 0.01) in the prefrontal [BA 9: (–22 26 40) (12
40 20), BA 10: (8 50 10)] and supplemental motor area {BA
8: (–8 32 46) (4 42 46)]. Significantly greater contralateral
activity (p < 0.01) was observed in the anterior cingulate
cortex [BA 32: (10 40 16)]. Controlling for depressed mood
(BDI) and resting pain (MPQ) did not eliminate the differ-
ences, but only the contralateral areas remained significant
(BA 9, BA 10, and BA 32).
Do FM patients exhibit greater neural responses to an
absolute pain stimulus? (Condition #3 or #4): perceptual
ratings. There were no significant differences between
groups for either pain intensity or unpleasantness ratings.
The average pain rating for the five 47°C stimuli was 2.2 ±
1.5 for the controls and 2.4 ± 2.1 for the FM group. Pain
unpleasantness ratings for the 5 stimuli were 1.8 ± 1.8 for
the controls and 2.7 ± 2.5 for the FM group.
Within-group analyses. Activation following five 47°C
stimuli in FM patients and controls is illustrated in Figure 6.
Areas of significant signal increase for the control group
occurred predominantly in the contralateral hemisphere
including the supplemental motor area (BA 6), sensory
cortex (BA 2), and the cingulate cortex (BA 24). Bilateral
activity was observed in the prefrontal cortex (BA 9), the
putamen of the lenticular nucleus, and the medial dorsal
nucleus of the thalamus.
The FM group exhibited robust activity mostly in the
contralateral hemisphere (Figure 6); however, several areas
showed bilateral activity. Bilateral activity was found in the
primary motor cortex (BA 4), supplemental motor area (BA
6), prefrontal cortex (BA 9), inferior parietal cortex (BA40),
the anterior cingulate cortex (BA 24 and BA 32), and the
caudate nucleus. Contralateral activation occurred in the
sensory cortex (BA 2 and BA 3), the supplemental motor
area (BA 8), the anterior and posterior insular cortex, the
putamen and globus pallidus of the lentiform nucleus, and
the thalamus including the ventroposterolateral nucleus and
the pulvinar.
The Journal of Rheumatology 2004; 31:2
370
Figure 3. Mean pain intensity (± SE) for the 8 random heat stimuli in
patients with FM (n = 9) and controls (n = 9).
Figure 4. Number of significant a priori pain-related sites as a function of
experimental condition for FM patients (n = 8) and healthy controls (n =
8). The figure shows the differences in pain-related brain regions across all
experimental conditions. For each condition the areas of activation are
based on within-subject’s group analyses with significance set at p < 0.01.
With the exception of the perceptually equivalent condition (pain rated as
5), FM subjects exhibited more pain-relevant brain areas in response to
both nonpainful and painful heat stimuli (*p = 0.05, Fisher’s exact test).
Between-group analyses. None of the areas observed in the
control group was found to be significantly greater than the
FM group. For the FM group, significantly greater activity
(p < 0.01) occurred in the contralateral insular cortex ([38
4–6]; Figure 7) compared to controls. Covariance for BDI or
MPQ totals did not eliminate the difference.
Do FM patients exhibit greater neural responses to a
perceptually equivalent pain stimulus? (Condition #3 or
#4): perceptual ratings. The mean temperature rated as “5”
for the control group was 48.3 ± 0.6°C. The average temper-
ature for the FM group was 47.4 ± 1.4°C. The difference of
almost 1° Celsius was large (Cohen d = 0.93). Moreover, the
range of temperatures for the FM group was 45°C to 48°C,
while the range for the control group was 47.5°C to 49°C.
There were no significant differences between groups for
self-reported pain intensity or unpleasantness ratings. The
average pain rating for the 5 stimuli presented during fMRI
scanning was 3.4 ± 1.3 for controls and 3.9 ± 2.6 for the FM
group. Pain unpleasantness ratings were 2.8 ± 1.5 for
controls and 3.2 ± 2.0 for the FM group.
Within-group analyses. Areas of significant activity for
controls occurred bilaterally in the anterior insular cortex
(Figure 8). Contralateral activity occurred in the supple-
mental motor area (BA 6), the prefrontal cortex (BA 9), and
the inferior parietal cortex (BA 40). Ipsilateral activity was
observed for the cingulate gyrus.
For the FM group, bilateral activity was observed in the
inferior parietal (BA 40) and anterior insular cortices
(Figure 8), although the insular activity was left lateralized
(65 voxels on the left vs 6 voxels on the right). Ipsilateral
activity was also observed in the prefrontal cortices (BA 9
and BA 10).
Between-group analyses. There were no significant differ-
ences in any brain areas for either FM or control group
comparisons. Thus when a perceptually equivalent pain
stimulus was delivered that consisted of greater absolute
temperatures for the control group, there were no significant
group differences in brain responses.
Do neural responses to nonpainful stimuli remain elevated
following the presentation of several painful stimuli?
(Condition #5): perceptual ratings. Consistent with Run #2,
there were no significant differences in self-reported pain or
unpleasantness for the 5 random stimuli. The average pain
rating for the FM group was 0.4 (± 0.2) and for controls 0.04
(± 0.03). The average unpleasantness rating for the FM
group was 0.8 (± 0.4) and for the control group 0.3 (± 0.1).
The intensity ratings were a significant decrease from Run
#2 for controls (p < 0.05), but not for the FM group.
Within-group analyses. The control group exhibited bilateral
activity in the superior parietal cortex (BA 7) and the thal-
amus. Contralateral activity was also observed in the supe-
rior parietal cortex (BA 5), prefrontal (BA 10), anterior
insular and cingulate (BA 23, BA 24, BA 31, BA 32)
cortices. Although not an a priori region, the controls also
Cook, et al: Functional imaging of FM pain 371
Figure 5. A. Relationship between the blood oxygen level-dependent (BOLD) activity and the experimental paradigm during practice in a
representative patient with FM. The figure depicts the predicted hemodynamic adjusted response based on the experimental paradigm (solid
lines), the actual response (broken line), and the superimposed block paradigm (bottom of the graph). (B) Movement parameters (x, y, z, roll,
pitch, and yaw) for the practice session for the same patient. The figure depicts the movement of one subject in the x, y, and z directions (top
panel) and the roll, pitch, and yaw directions (middle), as they related to the experimental block paradigm (bottom). Measured movement para-
meters were unrelated to either the experimental block paradigm or the BOLD response during the practice session.
exhibited significant activity within the periaqueductal
gray (PAG) of the midbrain (Figure 9).
The FM group exhibited bilateral activity in the supple-
mental motor areas (BA 6, BA 8), prefrontal (BA 9), supe-
rior parietal (BA 7), and anterior and posterior insular
cortices. Bilateral activation was also observed in the
ventral (left and right) and medial dorsal (left) nuclei of the
thalamus. Contralateral activity was observed in the
caudate and putamen. Ipsilateral activity was observed in
the anterior cingulate cortex (BA 24 and BA 32).
Between-group analyses. Controls showed significantly
greater activity (p < 0.01) in the contralateral medial dorsal
The Journal of Rheumatology 2004; 31:2
372
Figure 6. Pain-related activation in response to a 47°C stimulus in female
controls (n = 8) and FM subjects (n = 8). Significant increases in fMRI
signal are superimposed on a normalized T1 weighted brain and are based
on a voxel-based whole-brain analysis of each group separately. FM
patients exhibited a greater number of activated regions, and within similar
regions a greater number of activated pixels.
Figure 7. The greater BOLD signal in the anterior insular cortex of FM
patients compared to controls. Significant increases in fMRI signal are
superimposed on a normalized T1 weighted brain. Significance (p < 0.01)
is based on a ROI analysis of a specific region of the anterior insula that
exhibited significant activation in the within-group whole-brain search.
Cook, et al: Functional imaging of FM pain 373
Figure 8. Pain-related activation in response to a perceptually
equivalent pain stimulus (rated as 5) in controls (n = 8) and
patients with FM (n = 8). Significant increases in fMRI signal
are superimposed on a normalized T1 weighted brain and are
based on a voxel-based whole-brain analysis of each group
separately. When a perceptually equivalent stimulus was
presented that was of an absolute lower stimulus in FM patients,
qualitative differences in area and extent of activation were
diminished.
Figure 9. Periaqueductal gray (PAG) activation following several painful stimuli and in response to random warm stimuli in control subjects.
nucleus of the thalamus and the contralateral anterior cingu-
late cortex (BA 23: [5 –27 32] and BA 24: [5 –20 36]).
ANCOVA with BDI total as the covariate resulted in only
the contralateral anterior cingulate cortex (BA 23 and BA
24) activation remaining significantly greater than the FM
group. Similar results were obtained with the MPQ total as
the covariate. The FM group exhibited significantly greater
activity (p < 0.01) in the ipsilateral insula [32 14 12].
ANCOVA for either BDI or MPQ did not change the results.
Relationship between subjective pain ratings and brain
activity. Linear regression indicated that for both groups
combined, self-reported pain intensity and unpleasantness
were significantly (p < 0.01) related to bilateral activity in
the cingulate cortex, contralateral activity in the sensory
cortex (BA 2 and BA 3), and ipsilateral activity in the infe-
rior parietal (BA 40) and anterior insular cortices.
DISCUSSION
Consistent with our hypotheses and previous research indi-
cating dysregulation of the nociceptive system in FM57-10,28,
the FM group exhibited greater sensitivity to suprathreshold
experimental pain stimuli and greater responses in multiple
brain regions in response to both painful and nonpainful
stimuli. For nonpainful warm stimuli, the greatest differ-
ences were observed in the prefrontal cortex, the supple-
mental motor area, the insula, and the anterior cingulate
cortex. For painful heat stimuli, the greatest differences
were observed in the anterior insular cortex. These results
showing greater pain-related brain activity to both
nonpainful and painful stimuli in patients with FM
compared to controls provide support for a physiological
explanation for FM pain, and suggest that FM symptoms
may be maintained by amplified neural responses to afferent
sensory stimuli. Our investigation was also designed to
overcome several of the limitations of previous pain and
brain imaging research and to apply some of the controls
used in more recent brain imaging and pain studies14,28,29. In
particular, and with the noted exceptions14,28,29, brain
imaging investigations of pain have not consistently
controlled for attention, have not obtained concomitant pain
ratings while delivering the pain stimulus, and have not
measured pain in both absolute and relative terms.
Therefore, we instructed subjects to focus on a set of pain
scales during scanning and had them rate their pain intensity
just prior to removal of the stimulus. This allowed us to
examine more precisely the relationship between pain
ratings and the BOLD response to the stimulus. We also
examined pain in both perceptually equivalent and absolute
stimulus terms to determine the importance of pain percep-
tion in FM and to make more meaningful group compar-
isons. Finally, we examined both painful and nonpainful
stimuli to understand the function of the nociceptive system
in FM. The fMRI results revealing similar psychophysical
pain responses accompanied by greater neural responses in
the FM group suggest that FM (not just pain sensitivity) is a
disorder involving augmented physiological processing of
nociceptive stimuli.
Perceptual and neural responses to nonpainful stimuli. The
most striking feature of the results from the first set of warm
stimuli was the host of pain-related areas activated in the
FM group, but not the control group. Direct comparison
between groups indicated that FM subjects had significantly
greater activity in prefrontal, supplemental motor, and ante-
rior cingulate cortices, compared to controls. These regions
are well documented in pain and brain imaging research44,
and do not normally show significant activation when non-
noxious warm stimuli are delivered to healthy control
subjects13,15,16. Greater activity in the prefrontal and anterior
cingulate cortices suggests that cognitive and sensory
aspects of pain such as anticipation, attention, and pain
memories were greater in the FM group44. The role of the
supplemental motor cortex in pain processing is less clear,
but activity in this region may reflect motor inhibition to
avoid withdrawal of the stimulated hand. Importantly, these
differences were observed in the absence of any perceptual
differences in pain or unpleasantness ratings, and were not
eliminated when current depressed mood or pain was taken
into account. Thus, the presence of BOLD responses to
nonpainful stimuli occurring in regions of the brain
normally associated with processing of painful stimuli
provides evidence of augmented central responses to
sensory stimuli in FM, and supports the contention that FM
pain results from a dysregulation of the nociceptive system.
Asecond set of warm stimuli was given to determine the
recovery of neural responses following the presentation of a
series of painful stimuli. The results from the fMRI data
suggest that neither group recovered completely. Both
groups exhibited activity in pain-relevant regions. For the
controls, this represented a very different response from that
observed during the first set of warm stimuli, where only
thalamic activity reached significance. For the FM group,
the regions of activity were much the same as those
observed during the previous warm and pain runs. Group
comparisons indicated that the FM group had significantly
greater activity in the ipsilateral insular cortex, a region
consistently associated with pain perception44. Insular
cortex activity dramatically increases as a stimulus changes
from innocuous to noxious and is functionally implicated in
sensory intensity coding44. Thus, greater activity in this area
in the FM group indicates enhanced sensory coding of
nonpainful heat stimuli and is supportive of augmented
sensory processing. Moreover, the pattern of pain-relevant
fMRI responses to both painful and nonpainful stimuli, and
the significantly greater activity in brain areas involved in
both cognitive (prefrontal and cingulate cortices) and
sensory (supplemental motor, cingulate, and insular
cortices) regions, is evidence that CNS dysregulation in the
FM group was operating independent of stimulus condition.
The Journal of Rheumatology 2004; 31:2
374
The lack of thalamic activity in the FM group deserves
mention, as this result is consistent with PET and SPECT
studies in FM showing decreased thalamic blood flow22,25.
Additionally, a recent fMRI study28 described thalamic acti-
vation in response to pressure pain for controls but not
patients with FM. This result was interpreted as tonic inhi-
bition of thalamic nuclei in FM resulting from persistent
excitatory afferent pain signaling. This theoretical mecha-
nism is supported by results from other chronic pain condi-
tions, where thalamic blood flow was normalized upon
chronic pain relief24,25. Our results extend these findings by
showing a similar result in response to a nonpainful stim-
ulus. Another interesting observation during the final warm
run was the presence of PAG activity in the control group
but not the FM group. The PAG is well established as an
important area involved in the descending modulation of
nociceptive signals45. PET studies examining the cognitive
and affective modulation of pain have reported PAG
activity46-49. While not an a priori assumption, the absence
of PAG activity in the FM group suggests that pain modula-
tory processes were not occurring after receiving several
painful stimuli, while the pain regulatory system of the
control subjects was actively attempting to inhibit further
nociceptive input. These results are in agreement with inves-
tigations suggesting a dysregulation of noxious inhibitory
controls in FM5,7,8.
Perceptual and neural responses to painful stimuli. We gave
both groups an absolute stimulus (47°C) and a relative stim-
ulus (pain previously rated as 5) to examine whether differ-
ences observed between FM and control groups were due to
the temperature delivered or to the perception of pain. To
our surprise the FM group did not rate the 47°C stimulus as
significantly more painful than the controls. Therefore,
while Experiment 1 demonstrated that the FM subjects were
generally more sensitive to experimental pain stimuli, this
result did not generalize to the fMRI session. Areview of the
literature indicates that such a lack of differences is not
unusual in FM. Petzke, et al50 reported that FM patients
found equal-pressure pain to be less unpleasant than a group
of pain-free controls. Consistent with this result, our FM
groups rated the 10 random pain stimuli delivered in
Experiment 1 as more intense than controls but not more
unpleasant. Moreover, investigations using similar
psychophysical techniques have reported values consistent
with those from our investigation4,8.
In the absence of perceptual differences, the brain
responses to the painful stimulus were quite different
between the FM and control groups. When directly
compared to the control group, the greatest difference
occurred within the anterior insular cortex. The anterior
insular cortex is the most consistent region of activation
reported in pain and brain imaging studies44. That the ante-
rior insula is implicated in intensity coding, and that activity
within this region was greater in the FM group, is consistent
with the view that FM pain is the result of augmented
processing of nociceptive stimuli28,44. Specifically, the
results showing similar pain ratings to the absolute stimulus
but greater neural activity suggests that given the same
afferent information, the physiological response of the FM
patient is exaggerated. Additionally, qualitative differences
were apparent both in the number of regions activated and
in the number of pixels activated within a given region,
further implicating augmented nociceptive processing.
Thus, pain in FM may be due in part to an augmented
central reaction to incoming sensory stimuli that remains
elevated after the removal of the afferent signal. Evidence of
an exaggerated wind-up response in FM that can be attenu-
ated with an NMDA antagonist supports this contention and
further suggests that NMDA receptor sensitivity may act to
maintain CNS sensitivity in FM9,51.
As intended, ratings for the relative stimulus were not
different between the FM and control groups. However, for
the FM group the temperatures necessary to reach a pain
rating of 5 on the CR-10 scale were on average 1° Celsius
lower than controls (48.3°C for controls and 47.4°C for the
FM group). These temperatures represented a large change
from the absolute 47°C stimulus for controls (1.3°C
increase) and a small change for the FM group (0.4°C
increase). Moreover, the range of pain temperatures indi-
cated that patients and controls could differ by as much as
4°C, with a range of 45°C to 48°C in the FM group and a
range of 47.5°C to 49°C in the control group. As a result, the
controls showed an increase in activation that would be
expected when a stronger painful heat stimulus is deliv-
ered52, while the FM group exhibited fewer regions of acti-
vation, consistent with several subjects receiving an
absolutely lower stimulus, and a general change in laterality
from contralateral to ipsilateral dominance, the relevance of
which is not presently clear. Thus, when a perceptually
equivalent stimulus was delivered that was of an absolute
greater intensity for the controls, the FM group exhibited a
response similar to that of the control group. These results
for painful heat are consistent with those of Gracely and
colleagues28, who demonstrated similar neural responses in
FM patients and controls resulting from equal-pressure pain
intensity. They interpreted their findings as evidence of
central augmentation of pain in FM. Our results provide
additional evidence of augmented pain processing in
response to heat, and emphasize the need to compare
perceptually relevant stimuli when examining FM pain.
Neural responses to the anticipation of pain. To account for
potential brain activity associated with the movement
required to rate the stimuli using the handheld device and
the activity associated with the contact thermode touching
the skin, we employed a practice session, when subjects
practiced the rating task while not receiving any stimuli.
This session allowed participants to become accustomed to
the fMRI environment before receiving painful stimuli. As
Cook, et al: Functional imaging of FM pain 375
expected, the controls did not show any activity above base-
line in pain-relevant areas during this condition. This result
is consistent with data reported by Davis, et al19, using an
event-related paradigm, showing no pain-related activation
during a simulated ratings task. However, the FM group
exhibited activity in several brain areas previously identified
as pain-relevant that was not due to task-related movement.
These areas included frontal, parietal, and insular cortices
and suggest that the FM participants were anticipating a
painful stimulus53 or were in a heightened attentive state
while rating, even after being informed that no stimulus
would be delivered. Although we did not anticipate this
finding, it raises the possibility that FM participants were
more vigilant to the potential of incoming stimuli compared
to controls. Hypervigilance to experimental pain stimuli in
FM has been reported4-6. Research aimed at manipulating
vigilance and examining fMRI responses is needed to deter-
mine what role vigilance plays in the neural representation
of pain in FM.
Relationship between subjective pain ratings and brain
activity. Pain ratings for both groups were significantly and
positively related to increased activity in several pain-rele-
vant brain regions. This result is encouraging, given that in
many cases the self-report of pain does not correlate well
with physiological indicators of pain (e.g., muscle damage).
It is also in agreement with studies describing positive rela-
tionships between pain ratings and several brain regions,
including those observed in our investigation9,54.
Examination of a greater range of pain ratings may help to
further delineate the role of specific brain regions involved
in pain perception.
One potential limitation of this study is the small
between-group differences observed in pain perception
during fMRI testing. If we had observed greater differences
in pain perception our results might have been different in
magnitude but likely in the same direction as the current
data. The absence of perceptual differences accompanied by
an exaggerated neural response to the 47°C stimulus in FM
actually strengthens the notion that FM pain, not sensitivity
to an experimental stimulus, is a disorder of physiological
processing of afferent sensory information. Another poten-
tial limitation is the modest between-group psychophysical
differences observed in Experiment 1. While our results are
similar to those of previous investigations, others have
reported more robust psychophysical differences. Our
choice of a reaction time-inclusive method for measuring
pain threshold could be questioned. However, with a
syndrome such as FM we would expect pain thresholds to
be lower when using an ascending method of limits, since
this approach capitalizes on the subject’s anticipation of a
painful stimulus. It is also important that our FM group may
have comprised less severe cases than in previous investiga-
tions. For pilot study purposes, we screened our subjects to
ensure they were free from clinical depression and not
taking pain medications. This may have resulted in recruit-
ment of a relatively high-functioning FM group, in which
pain sensitivity was not inflated by comorbid depression.
However, all subjects did meet stringent diagnostic criteria
for FM1. Finally, the sample size of the study is a potential
limitation. However, the robust differences in the fMRI
experiment that remained even after controlling for potential
confounds indicates that power was not a problem.
The etiology of FM is unknown, and no consistent under-
lying mechanism explaining FM pain has been identified.
One theory that has emerged prominently in the literature
focuses on the mounting evidence indicating abnormalities
of the nociceptive system in patients with FM and attempts
to explain FM pain as a dysregulation of the physiological
processing of afferent sensory information. Our investiga-
tion was designed to test this theory by examining fMRI
brain responses to both painful and nonpainful stimuli and
examining pain in both absolute stimulus and perceptually-
relevant terms. Consistent with our hypotheses, patients
with FM exhibited neural activation in brain regions associ-
ated with pain perception in response to nonpainful stimuli,
with the greatest differences occurring in prefrontal, supple-
mental motor, insular, and cingulate cortices. Also consis-
tent with our hypotheses, patients with FM displayed greater
activity than controls to an absolute pain stimulus but
similar activity to a perceptually equivalent stimulus. The
difference was most apparent in the anterior insular cortex,
a region consistently associated with encoding of painful
stimuli. Our results indicate that brain responses to sensory
stimuli, in regions involved in sensory, cognitive, and
emotional aspects of the pain experience, are augmented in
FM. These results support a physiological explanation of
FM pain and provide objective evidence of cortical and
subcortical amplification of both painful and nonpainful
thermal stimuli.
REFERENCES
1. Wolfe F, Smythe HA, Yunus MB, et al. The American College of
Rheumatology 1990 criteria for the classification of fibromyalgia:
Report of the Multicenter Criteria Committee. Arthritis Rheum
1990;33:160-72.
2. Mikkelsson M, Pirjo MD, Kautiainen H, Isomeri R, Isomaki H.
Muscle and bone pressure pain threshold and pain tolerance in
fibromyalgia patients and controls. Arch Phys Med Rehabil
1992;73:814-8.
3. Gibson SJ, Littlejohn GO, Gorman MM, Helme RD, Granges G.
Altered heat pain thresholds and cerebral event-related potentials
following painful CO2laser stimulation in subjects with
fibromyalgia syndrome. Pain 1994;58:185-93.
4. Lautenbacher S, Rollman GB, McCain GA. Multi-method
assessment of experimental and clinical pain in patients with
fibromyalgia. Pain 1994;59:45-53.
5. Kosek E, Ekholm J, Hansson P. Modulation of pressure pain
thresholds during and following isometric contraction in patients
with fibromyalgia and in healthy controls. Pain 1996;64:415-23.
6. Sorensen J, Graven-Nielsen T, Henriksson KG, Bengtsson M,
Arendt-Nielsen L. Hyperexcitability in fibromyalgia. J Rheumatol
1998;25:152-5.
The Journal of Rheumatology 2004; 31:2
376
7. Kosek E, Hansson P. Modulatory influence on somatosensory
perception from vibration and heterotopic noxious conditioning
stimulation (HNCS) in fibromyalgia patients and healthy subjects.
Pain 1997;70:41-51.
8. Lautenbacher S, Rollman GB. Possible deficiencies of pain
modulation in fibromyalgia. Clin J Pain 1997;13:189-96.
9. Staud R, Vierck CJ, Cannon RL, Mauderli AP, Price DD. Abnormal
sensitization and temporal summation of second pain (wind-up) in
patients with fibromyalgia syndrome. Pain 2001;91:165-75.
10. Mengshoel AM, Saugen E, Forre O, Vollestad NK. Muscle fatigue
in early fibromyalgia. J Rheumatol 1995;22:143-50.
11.Talbot JD, Marrett S, Evans AC, Meyer E, Bushnell MC, Duncan
GH. Multiple representations of pain in human cerebral cortex.
Science 1991;251:1355-8.
12. Coghill RC, Talbot JD, Evans AC, et al. Distributed processing of
pain and vibration by the human brain. J Neurosci
1994;14:4095-108.
13. Casey KL, Minoshima S, Morrow TJ, Koeppe RA. Comparison of
human cerebral activation pattern during cutaneous warmth, heat
pain, and deep cold pain. J Neurophysiol 1996;76:571-81.
14. Davis KD, Taylor SJ, Crawley AP, Wood ML, Mikulis DJ.
Functional MRI of pain- and attention-related activations in the
human cingulate cortex. J Neurophysiol 1997;77:3370-80.
15. Derbyshire SWG, Jones KPJ, Gyulai F, Clark S, Townsend D,
Firestone LL. Pain processing during three levels of noxious
stimulation produces differential patterns of central activity. Pain
1997;73:431-45.
16. Davis KD, Kwan CL, Crawley AP, Mikulis DJ. Functional MRI
study of thalamic and cortical activations evoked by cutaneous heat,
cold, and tactile stimuli. J Neurophysiol 1998A;80:1533-46.
17. Casey KL, Minoshima S, Berger KL, Koeppe RA, Morrow TJ, Frey
KA. Positron emission tomographic analysis of cerebral structures
activated specifically by repetitive noxious heat stimuli.
J Neurophysiol 1994;71:802-7.
18. Rainville P, Duncan GH, Price DD, Carrier B, Bushnell C. Pain
affect encoded in human anterior cingulate but not somatosensory
cortex. Science 1997;277:968-71.
19. Davis KD, Kwan CL, Crawley AP, Mikulis DJ. Event-related fMRI
of pain: entering a new era in imaging pain. Neuroreport
1998B;9:3019-23.
20. Peyron R, Larrea-Garcia L, Gregoire MC, et al. Haemodynamic
brain responses to acute pain in humans: sensory and attentional
networks. Brain 1999;122:1765-79.
21. Ploghaus A, Tracey I, Gati JS, et al. Dissociating pain from its
anticipation in the human brain. Science 1999;284:1979-81.
22. Mountz JM, Bradley LA, Modell JG, et al. Fibromyalgia in women:
Abnormalities of regional cerebral blood flow in the thalamus and
the caudate nucleus are associated with low pain threshold levels.
Arthritis Rheum 1995;38:926-38.
23. Kwiatek R, Barnden L, Tedman R, Jarrett R, Chew J, Rowe C, Pile
K. Regional cerebral blood flow in fibromyalgia — Single-photon-
emission computed tomography evidence of reduction in the pontine
tegmentum and thalami. Arthritis Rheum 2000;43:2823-33.
24. Di Piero V, Jones AKP, Iannotti F, et al. Chronic pain: a PET study
of the central effects of percutaneous high cervical cordotomy. Pain
1991;46:9-12.
25. Hsieh J, Belfrage M, Stone-Elander S, Hansson P, Ingvar M. Central
representation of chronic ongoing neuropathic pain studied by
positron emission tomography. Pain 1995;63:225-36.
26. Jones AKP, Derbyshire SWG. Reduced cortical responses to noxious
heat in patients with rheumatoid arthritis. Ann Rheum Dis
1997;56:601-7.
27. Fukumoto M, Ushida T, Zinchuk VS, Yamamoto H, Yoshida S.
Contralateral thalamic perfusion in patients with reflex sympathetic
dystrophy syndrome. Lancet 1999;354:1790-1.
28. Gracely, RH, Petzke F, Wolf JM, Clauw DJ. Functional magnetic
resonance imaging evidence of augmented pain processing in
fibromyalgia. Arthritis Rheum 2002;46:1333-43.
29. Fukuda K, Straus SE, Hickie I, et al. The chronic fatigue syndrome:
Acomprehensive approach to its definition and study. Ann Intern
Med 1994;121:953-9.
30. Marcus S, Robins LN, Bucholz K. Quick diagnostic interview
schedule 3R, version 1. St. Louis, MO: Washington University
School of Medicine; 1990.
31. Spielberger CB. State-trait anxiety inventory. Palo Alto: Counseling
Psychologists Press; 1983.
32. Beck AT, Ward CM, Mendelsohn M, Mock J, Erbaugh M. An
inventory for measuring depression. Arch Gen Psychiatry
1961;5:561-71.
33. Kohn PM. Sensation seeking, augmenting reducing and strength of
the nervous system. In: Spence JT, Izard C, editors. Motivation,
emotion and personality. Proceedings, XXIII Congress of
Psychology, Acapulco, Mexico, September 2-7, 1984. Amsterdam:
Elsevier; 1985:167-73.
34. Melzack R. The short-form McGill Pain Questionnaire. Pain
1987;30:191-7.
35. McDermid AJ, Rollman GB, McCain GA. Generalized
hypervigilance in fibromyalgia: Evidence of perceptual
amplification. Pain 1996;66:133-44.
36. Cook DB, O’Connor PJ, Eubanks SA, Smith JC, Lee M. Naturally
occurring muscle pain during exercise: assessment and experimental
evidence. Med Sci Sports Exerc 1997;29:999-1012.
37. Borg G. A category scale with ratio properties for intermodal and
interindividual comparisons. In: Geisler HG, Petzold P, editors.
Psychophysical judgment and the process of perception. Berlin:
VEB Deutscher Verlag der Wissenschaften; 1982:25-34.
38. Harms-Ringdahl K, Brodin H, Eklund L, Borg G. Discomfort and
pain from loaded passive joint structures. Scand J Rehabil Med
1983;15:205-11.
39. Marks LE, Borg G, Ljunggren G. Individual differences in
perceived exertion assessed by two new methods. Percept
Psychophys 1983;34:280-8.
40. Neely G, Ljunggren G, Sylven C, Borg G. Comparison between the
visual analogue scale (VAS) and the category ratio scale (CR-10) for
the evaluation of leg exertion. Intern J Sports Med 1992;13:133-6.
41. Borg G. The Borg CR10 scale. In: Borg’s perceived exertion and
pain scales. Champaign: Human Kinetics; 1998:39-43.
42. Friston KJ, Holmes AP, Poline JB, et al. Analysis of fMRI
time-series revisited. Neuroimage 1995;2:173-81.
43. Cohen J. Statistical power analysis for the behavioral sciences. 2nd
ed. Hillsdale, NJ: Erlbaum Associates; 1988.
44. Peyron R, Laurent B, Garcia-Larrea L. Functional imaging of brain
responses to pain. A review and meta-analysis (2000). Neurophysiol
Clin 2000;30:263-88.
45. Fields HL, Basbaum AI. Central nervous system mechanisms of
pain modulation. Textbook of pain. 3rd ed. New York: Churchill
Livingstone; 1994:2443-57.
46. Hsieh JC, Stone-Elander S, Ingvar M. Anticipatory coping of pain
expressed in the human anterior cingulated cortex: a positron
emission tomography study. Neurosci Lett 1999;262:61-4.
47. Petrovic P, Ingvar I. Imaging cognitive modulation of pain
processing. Pain 2002;95:1-5.
48. Petrovic P, Petersson KM, Ghatan PH, Stone-Elander S, Ingvar M.
Pain-related cerebral activation is altered by a distracting cognitive
task. Pain 2000;85:19-30.
49. Tolle TR, Kaufmann T, Siessmeier T, et al. Region-specific
encoding of sensory and affective components of pain in the human
brain: a positron emission tomography correlation analysis. Ann
Neurol 1999;45:40-7.
50. Petzke F, Clauw DJ, Benson E, Gracely RH. Unpleasantness of
Cook, et al: Functional imaging of FM pain 377
The Journal of Rheumatology 2004; 31:2
378
induced pressure pain in fibromyalgia patients and healthy controls
[abstract]. Arthritis Rheum 2000;43 Suppl:S210.
51. Graven-Nielsen T, Kendall SA, Henriksson KG, et al. Ketamine
reduces muscle pain, temporal summation, and referred pain in
fibromyalgia patients. Pain 2000;85:483-91.
52. Coghill RC, Sang CN, Maisog JM, Iadarola MJ. Pain intensity
processing within the human brain: a bilateral, distributed
mechanism. J Neurophysiol 1999;82:1934-43.
53. Sawamoto N, Honda M, Okada T, et al. Expectation of pain
enhances responses to nonpainful somatosensory stimulation in the
anterior cingulate cortex and parietal operculum/posterior insula: an
event-related functional magnetic resonance imaging study.
J Neurosci 2000;20:7438-45.
54. Schneider F, Habel U, Holthusen H, et al. Subjective ratings of pain
correlate with subcortical-limbic blood flow: an fMRI study.
Neuropsychobiol 2001;43:175-85.
APPENDIX
Pain Threshold Methods for Experiment 1
Prior to threshold measurement, participants were given a standard set of
instructions for both heat and pain threshold assessments. For heat
threshold, participants were asked to indicate, by pressing a mouse button
with their free hand, the point at which they felt any change in temperature.
For pain threshold, participants were asked to indicate when the warm
sensation became “just painful” and were reminded that this was not a test
to see how much pain they could tolerate. Three sets of heat stimuli
followed by 3 sets of pain stimuli were delivered separated by 3 minutes
each. Each set consisted of 3 threshold measurements. Heat and pain
thresholds were defined as the average temperature of sets 2 and 3.
Suprathreshold Pain Methods for Experiment 1
During testing, the subject could see the pain scales using a set of fiber
optic goggles (Avotec, Jensen Beach, FL, USA). The goggles were inter-
faced with a laptop computer (Gateway 2000) for display. To obtain ratings
during the painful stimulus without the subject’s verbal response, ratings
were made using a keypad. The keypad was placed in the participant’s right
hand for recording pain scores and a Labview program (National
Instruments, Austin, TX, USA) allowed the subject to scroll up or down the
pain scale until the appropriate number representing their pain perception
was selected. Following a standard set of instructions describing the
purpose and use of the pain intensity and unpleasantness scales, the subject
reclined on an examination table and the Peltier thermode was attached to
the hand. Subjects were told that they were about to receive a number of
stimuli, some of which may be “extremely painful,” but that no skin
damage would result. Subjects were prompted to record their ratings during
the last 5 s of each 10 s stimulus period.
Experimental Paradigm for fMRI Testing
Run 1 (Condition 1) was used to examine the potential activation associ-
ated with the thermode touching the skin (no stimulus given) and the small
movement required for rating with the right hand. Participants were told
that no stimulus would be delivered and were given a number to record
during the last 5 s of each on-period. Runs 2 and 5 (Conditions 2 and 5)
consisted of 5 random warm stimuli ranging from 34°C to 42°C in 2°C
increments. Participants all received the same random sequence and were
prompted during the last 5 s of each stimulus to rate the pain intensity. Runs
3 and 4 consisted of painful heat stimuli. One run consisted of five 47°C
stimuli and the other consisted of a temperature that was previously rated
as a 5 (“strong pain”) on the 0–10 pain intensity scale in Experiment 1. The
runs were counterbalanced so that half the subjects received the absolute
stimulus (47°C) first and half received the relative stimulus (temperature
previously rated as 5; the verbal anchor “strong pain”) first. The separate
presentation of absolute (47°C) and relative (temperature previously rated
5) stimuli was intended to examine if fMRI brain responses to painful heat
are different when the stimuli are: (a) objectively identical but perceptually
distinct (absolute 47°C stimulus) versus (b) perceptually identical but
objectively distinct (stimulus previously rated as 5, “strong pain”).
Immediately after each run, participants rated their peak pain unpleasant-
ness (0–10).
