w w w . e l s e v i e r . c o m / l o c a t e / p a i n

PAIN� 145 (2009) 18–23

Research papers

Pain relief through expectation supersedes descending inhibitory deficitsin fibromyalgia patients

Philippe Goffaux, Juliana Barcellos de Souza, Stéphane Potvin, Serge Marchand *

Université de Sherbrooke, Faculté de Médecine, Neurochirurgie, 3001, 12e Avenue Nord, Sherbrooke, Qué., Canada J1H 5N4

a r t i c l e i n f o a b s t r a c t

Article history:Received 31 July 2008Received in revised form 8 January 2009Accepted 4 February 2009

Keywords:FibromyalgiaExpectationsWithdrawal reflexSpineBrainDescending inhibition

0304-3959/$36.00 � 2009 International Associationdoi:10.1016/j.pain.2009.02.008

* Corresponding author. Tel.: +819 346 1110x1588E-mail address: Serge.Marchand@USherbrooke.ca

In healthy adults, expectations can modulate the activity of inhibitory bulbo-spinal projections, and caneven block the analgesic properties of counter-irritation – a phenomenon that triggers descending inhi-bition. Since descending inhibition is known to be deficient in fibromyalgia (FM) patients, we tested thepossibility that expectancy-mediated analgesia would improve, or even kick-start, the deficient inhibi-tory responses of FM patients. By measuring subjective pain ratings, spinal withdrawal reflexes, andsomatosensory evoked potentials (SEP), it was possible to test whether or not expectancy-mediated anal-gesia involved descending inhibition in FM patients. Here, we show that expectations of analgesia radi-cally change the subjective experience of pain, but do not eliminate evidence of spinal hyperexcitabilityin FM patients. We found that expectations of analgesia reduce subjective pain ratings and decrease SEPamplitudes, confirming that expectations influence thalamocortical processes. However, even when anal-gesia was experienced, the spinal activity of FM patients was abnormal, showing heightened reflexresponses. This demonstrates that, unlike healthy subjects, the modulation of pain by expectations inFM fails to influence spinal activity. These results indicate that FMs are capable of expectancy-inducedanalgesia but that, for them, this form of analgesia does not depend on the recruitment of descendinginhibitory projections.

� 2009 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.

1. Introduction

Contemporary neuroscientific thinking posits that our interpre-tation of sensory events is construed by our subjective experienceand depends on the activity taking place in our brain [2,17,26]. Acommon example of how sensory events can be modified by sub-jective states is placebo analgesia. Recent research confirms thatplacebo effects, or expectation effects, depend on the activityoccurring within the prefrontal cortex, the anterior cingulate, andthe periaqueductal grey [25,30]. Expectations of relief also affectthe strength of spinal nociceptive activity, thus blocking pain atthe earliest levels of the central nervous system [15,22]. Thismeans that top-down information can affect bottom-up signals ina way that is consistent with what is anticipated or contextuallyappropriate.

In an original study on this issue, we showed that spinal painprocessing is altered because expectations change the strength ofdescending pain inhibitory systems [15]. Descending systems orig-inate in the midbrain and respond to incoming pain signals bysending inhibitory projections throughout the spine [13,21].Importantly, descending inhibitory responses increase when painrelief is expected and are completely blocked when more pain is

for the Study of Pain. Published by

9; fax: +819 564 5424.(S. Marchand).

expected [15], resulting in spinal changes that have neuropathic-like features.

To appreciate the clinical benefits of expectations, we tested pa-tients suffering from fibromyalgia (FM). FM patients were selectedbecause they are known to suffer from intractable chronic pain[33], and have documented descending inhibitory deficits[16,20]. The objective, therefore, was to test the possibility thatexpectations of analgesia would improve, or even kick-start, thedescending inhibitory responses of FM patients. To explore this,we used a human model of electrical pain and descending inhibi-tion. Pain was evoked through transcutaneous electrical stimula-tions of the sural nerve, whereas descending inhibition wasactivated by using a standard counter-irritation procedure (immer-sion of the arm in a bath of painfully cold water). Half of the par-ticipants expected that the immersion procedure would alleviatetheir sural nerve pain sensations (analgesia group), and the otherhalf expected that it would increase their pain (hyperalgesiagroup). Groups were compared on immersion-induced changes inperceived sural nerve pain and on the amplitude of the withdrawalreflex response triggered by each sural nerve stimulation. Somato-sensory-evoked potentials (SEP) – measures of cortical response tosural nerve stimulation – were also recorded and compared be-tween groups. Therefore, expectation-induced changes occurringat spinal (reflex response) and supraspinal levels (perceived painand SEP responses) were measured. If expectations of analgesia

Elsevier B.V. All rights reserved.

P. Goffaux et al. / PAIN� 145 (2009) 18–23 19

help FM improve their descending inhibitory responses, then theimmersion procedure will decrease the intensity of nociceptive sig-nals at the earliest stages of nociceptive processing in the spinalcord. As a result, markers of spinal (and cortical) activity will de-crease in amplitude. If, however, expectations of analgesia only af-fect pain perception and fail to interact with descending circuits,then the immersion procedure will result in a reduction of per-ceived pain and SEP, but spinal nociceptive reflexes will remainlarge despite subjective reports of analgesia. In this situation, acontinued deficit in descending spinal inhibition would indicate arobust deficit in the normal functioning of inhibitory circuits(unrelated to expectation), and would confirm a fundamentalchange in the underlying neurobiology of expectation effects whenchronic pain is present.

2. Materials and methods

2.1. Subjects

Fourteen women suffering from FM were included in this study.All patients were diagnosed using the American College of Rhuma-tology’s classification criteria [32] and tested at the Sherbooke Uni-versity Hospital Center. Patients were asked to continue theircurrent medical treatment. However, if medication was necessaryfor breakthrough pain, testing was re-scheduled (see Table 1 fora list of medications used by patients). All patients signed a con-sent form approved by the local Ethics Committee.

2.2. Pain ratings

Pain was evaluated using two numeric scales ranging from 0 (nopain) to 100 (most intense/unpleasant pain imaginable). In ourstudy, pain intensity and unpleasantness ratings did not differ(all Z < 1.80, all Ps > .07), therefore, only the results for pain inten-sity are reported and discussed.

2.3. Sural nerve stimulations

The sural nerve was stimulated over its retromalleolar pathevery 7 s. Stimulations consisted of a volley of five electrical pulses(square waves – each 1 ms long) administered at a rate of 240 Hzusing a constant current stimulator. A stimulation volley lastedonly 21 ms. Stimulations produced a withdrawal reflex that wasmeasured via an electromyographic (EMG) recording of the kneeflexor muscle. Stimulations were set at 15% above the thresholdvalue required to provoke the reflex. Threshold values were deter-mined using an iterative staircase method [31]. EMG activity wasconsidered reflexive when its amplitude exceeded baseline activitylevels by at least 1.5 standard deviations. Withdrawal reflex activ-ity was quantified by calculating the integral of the rectified EMGsignal between 90 and 150 ms post-stimulation.

Table 1Use of analgesic/antihyperalgesic medication in fibromyalgia patients (sum of casesobserved).

Analgesic/antihyperalgesic medication Analgesia group(N = 7)

Hyperalgesia group(N = 7)

Tricyclic antidepressants 2 1Serotonin specific reuptake inhibitors 1 1Serotonin-norepinephrine reuptake

inhibitors1 2

Opiates 0 2Antiepileptic drugs (including calcium

channel blockers)1 2

Non-steroidal anti-inflammatory drugs 1 1Acetaminophen 2 2

2.4. Somatosensory-evoked brain potentials

SEPs were time-locked to each sural nerve stimulation and wereobtained at the vertex (Cz) using a monopolar montage with aright-ear reference. Using a Powerlab system (ADInstruments),our signal was sampled at 200 Hz and filtered online within a0.5–30 Hz bandpass. Electrooculographic (EOG) activity was alsorecorded and EOG artifacts (±60 lV) were eliminated from theanalyses. SEPs were baseline corrected between 0 and 100 ms priorto stimulation onset and averaged together. SEP responses wereaveraged across 17 trials during the immersion period and 34 trialsduring the pre-immersion period. From the mean waveforms, spe-cific epochs were isolated in order to capture both early (P45,N100) and late (N150, P260) somatosensory components. P45activity was examined in the 45–55 ms post-stimulus intervalwhereas N100 activity was examined in the 90–120 ms interval.N150 activity was examined in the 135–150 ms interval andP260 activity was examined in the 280–350 ms interval. It isimportant to point out that peak-scoring of the P260 is not advisedwhen using electrical stimulations, since, during this time (andslightly earlier), there is a spatial and temporal overlap of the scalpelectric potentials generated by both innocuous and pain-relatedsources [9]. In contrast, the 280–350 ms interval of this late poten-tial has been shown to reflect pain-related activity more specifi-cally [4,5].

2.5. Pre-experimental procedure

Prior to testing, participants were familiarized with the armimmersion procedure and the temperature of the water was ad-justed individually to insure that it produced strong but tolerablepain in all subjects (pain intensity P 30). After adjustment, theaverage bath temperature was 13.5 �C (±0.82) for the analgesiagroup and 13.5 �C (±0.77) for the hyperalgesia group. These tem-peratures are warmer than those typically used to trigger inhibi-tory responses in healthy adults (usually between 1 and 7 �C).However, pilot testing confirmed that temperatures well below13 �C were not well tolerated by patients. Furthermore, our previ-ous research demonstrates that temperatures in the 12–13 �Crange will produce strong analgesic responses, even in healthyadults [29]. As a result, we are confident that our immersion pro-cedure was well suited to maximise the inhibitory responses ofFM patients while also reducing the risk of attrition (owing to anexcessively painful immersion session). Following the pre-testingimmersion period, participants were fitted with electromyographicand electroencephalographic recording electrodes. Set-up lastedapproximately 40 min. This delay was important because it helpedreduce any potential carry-over effects from the pre-testingimmersion session to the experimental test.

Once set-up was completed, participants rated the extent towhich they expected their sural nerve pain to change during theimmersion procedure. Expectation ratings were given as a percent-age increase (+) or decrease (�) in pain. Patients were given noprior information regarding the putative effects of the immersionprocedure (i.e., expectations were not manipulated). We neverfound an individual who had no expectations, or who expectedthat the immersion procedure would have no effect. As a result,it was impossible to include a control group where expectationswere neutral or absent.

2.6. Experimental protocol

Testing began with a 5-min baseline condition during whichsubjects remained seated and relaxed. Sural nerve stimulationswere then administered repeatedly for 10 min. Each stimulationprovoked a withdrawal reflex and a somatosensory evoked brain

-50

0

50

Cha

nge

in P

ain

Inte

nsity

(%

rela

tive

to b

asel

ine)

ExpectedExperienced

a

b

Analgesia

Hyperalgesia

Fig. 1. (a) Schematic of the testing procedure, including the 2-min immersionperiod. Testing lasted 10 min in total. (b) Fibromyalgia patients who expectedanalgesia perceived large reductions (negative values) in sural nerve pain duringthe immersion period, whereas fibromyalgia patients who expected hyperalgesiaperceived large increases (positive values) during the immersion period. Scores areexpressed as a percentage value from baseline (i.e., the period of time prior toimmersion).

20 P. Goffaux et al. / PAIN� 145 (2009) 18–23

potential. Four minutes after sural nerve stimulation began, partic-ipants immersed their right arm in cold water for 2 min (Fig. 1a).Pain ratings pertaining to sural nerve stimulations were recordedevery minute.

2.7. Statistical analyses

All data are given as means ± SD. Comparisons made betweengroups were conducted using the Mann–Whitney test, a non-para-metric test appropriate for small samples. Within group analyseswere conducted using the Wilcoxon test and one-sample t-tests.To further reduce the risk of interpreting empirically over-fitted re-sults (arising from a relatively small number of subjects per group),and to better appreciate the significance of our data, a measure ofeffect size (g2) was provided. Relationships between our differentoutcome variables were looked at using Spearman correlations.P < .05 (two-tailed) was considered statistically significant.

3. Results

3.1. Pre-experimental parameters

The analgesia group (N = 7) and the hyperalgesia group (N = 7)had similar withdrawal reflex thresholds (11.9 mA ±0.7 for theanalgesia group and 11.9 mA ±1.4 for the hyperalgesia group;U = 24.50, P = 1.0, g2 = 0.01) and perceived pre-immersion suralnerve stimulations as comparably painful (mean = 19.3 ± 3.4 forthe analgesia group, and, mean = 29.5 ± 5.9 for the hyperalgesiagroup; U = 16.00, P = .32, g2 = 0.16), suggesting that our groupswere well matched on baseline pain sensitivity. Patients were also

well matched in terms of age. FM participants in the analgesiagroup were 53.3 (±2.8) years of age while FM participants in thehyperalgesia group were 47.3 (±4.3) years of age (U = 15.00,P = .22, g2 = 0.10).

3.2. Experimental findings

Results showed that the two groups expected and experiencedlarge differences in the effect of the arm immersion on sural nervepain intensity (both Us < 1.00, both Ps < .001, both g2 > 0.59;Fig. 1b). Whereas the analgesia group experienced a 24% decreasein perceived sural nerve pain during the immersion procedure, thehyperalgesia group experienced an 18% increase in their pain.Group differences in subjective pain ratings were mirrored by cor-responding changes in the amplitude of SEP responses to the suralnerve shock (Fig. 2a and c). Notably, during the immersion period,the analgesia group showed a significant reduction in the ampli-tude of the early P45 somatosensory component (Z = 2.37, P = .02,g2 = 0.59) and the frontally mediated P260 component (Z = 2.36,P = .02, g2 = 0.84), whereas the hyperalgesia group showed a signif-icant increase in P45 amplitude (Z = 2.20, P = .03, g2 = 0.65). No sig-nificant reductions in amplitude were found for the hyperalgesiagroup (all Z < 1.35, all Ps > .18, all g2 < 0.26). Changes in perceivedsural nerve pain, P45, and P260 amplitude were positively corre-lated with expectation ratings (all rs > .54, all Ps < .03), suggestingthat expectations predict corresponding variations in pain inten-sity and neuroelectric brain activity (Fig. 3a–c).

During the immersion period, the amplitude of the withdrawalreflex response was significantly increased for the hyperalgesiagroup (48% increase, t = 2.40, P = .05). Surprisingly, this was alsoobserved for the analgesia group (44% increase, t = 3.02, P = .02),indicating that FM patients show evidence of spinal hyperexcit-ability even when analgesia is expected and experienced (Fig. 2band d). Changes in reflex amplitude did not correlate with expecta-tion ratings (r = .13, P = .67; Fig. 3d). For FM patients, therefore,spinal hyperexcitability is ubiquitous and cannot be attributed toexpectation effects.

4. Discussion

This study provides important insights into the neuronal mech-anisms underlying nociceptive processing in FM patients. Resultsindicate that FM patients are capable of expectancy-mediatedanalgesia but that, despite this, their descending inhibitory circuitsremain impaired. Instead of showing evidence of spinal inhibition,FM patients showed enhanced reflex responses during the immer-sion procedure, even when analgesia was expected and experi-enced. It is clear, therefore, that the spinal neurons of FMpatients have undergone permanent changes, and now show facil-itatory firing when heterosegmental afferents enter the spine. Par-adoxical spinal hyperexcitability in chronic pain patients has alsobeen reported in a recent study conducted by Coffin et al. [3]. Intheir study, Coffin et al. [3] found that rectal distensions, a nor-mally analgesic procedure, induced facilitation of the withdrawalreflex response in patients suffering from irritable bowel syndrome(IBS). In healthy adults, rectal distensions produce a progressiveinhibition of the cutaneomuscular flexion reflex that is propor-tional to the volume of distension. Since spinal hyperexcitability(not spinal inhibition) was observed in IBS patients, these resultsconfirm that some chronic pain patients express hyperexcitableneurons in the lumbosacral segments of their spinal cord. Our find-ings add to these results by demonstrating that spinal hyperexcit-ability is not a consequence of anticipated pain.

Because the analgesia group reported analgesia despite showingspinal hyperexcitability, cognitive re-appraisal of ascending signalsmust have taken place to allow the experience of subjective pain

0

100

0 10

%

min

0

100

0 10

%

min

Analgesia Expected

Hyperalgesia Expected

a

c

b

d

Fig. 2. (a) (Analgesia expected group) and (c) (hyperalgesia expected group) show the average sural nerve evoked brain potentials recorded prior to and during armimmersion. Early (P45, N100) and late (N150, P260) brain components are shown. Our waveforms clearly demonstrate a large group difference in the suppression of the lateP260 component. Larger P260 reductions are observed when analgesia is expected (a) than when hyperalgesia is expected (c). The immersion procedure also affected the P45component in different ways, depending on expected relief. P45 amplitudes decreased during immersion when analgesia was expected (a) and increased when hyperalgesiawas expected (c). (b) (Analgesia expected group) and (d) (hyperalgesia expected group) show the average reflex amplitudes recorded for each sural nerve stimulation andexpressed as a percentage value from baseline (i.e., the period of time prior to immersion). Despite experiencing different effects from the immersion procedure, bothfibromyalgia groups showed increased reflex amplitudes during this period (arrows), suggesting the presence of spinal hyperexcitability and the absence of descendinginhibition.

P. Goffaux et al. / PAIN� 145 (2009) 18–23 21

relief. This interpretation implies that the brain plays an importantrole in predicting expectation effects, which is consistent with thechange in P260 amplitude observed during the immersion period(i.e., larger P260 reductions during expected pain relief). TheP260 waveform is sensitive to changes in pain intensity and expec-tation effects [1,6,7] and its sources are located within the suprac-allosal part of the anterior cingulate cortex [14], a region involvedin the integration of sensory and experiential events [24]. Activityin this region is thought to regulate conflict and realign internalbodily states with expected outcomes [27]. The change in P260amplitude observed here is consistent with the idea that the ACCtransformed the meaning of afferent spinal signals to facilitatethe expression of pain relief.

Our SEP results also showed greater reductions in P45 ampli-tude during expectations of analgesia. Sources for the P45 wave-form are located near the primary somatosensory cortex (SI), aregion critically involved in registering sensory information[8,18]. Recent studies indicate that SI activity increases when theanticipated threat value of somatosensory signals increases[10,17], suggesting that this part of the cortex is involved in theearly attentional processing of threat. With respect to chronic painpatients, it can be argued that such early processing is particularlyimportant since it helps to expedite escape behaviours and reducesthe risk of (re)injury. The expectancy-induced change in P45amplitude observed in the current study confirms that FM patientsengage in early attention monitoring when a change in pain is

a b

c d

50

100

150

200

250

0 100 200

Refle

x Am

plitu

de

(% re

lativ

e to

bas

elin

e)

-10

-5

0

5

10

0 100 200

Chan

ge in

P26

0 Am

plitu

de (µ

V)

0

100

200

0 100 200

Chan

ge in

Pai

n In

tens

ity

(% re

lativ

e to

bas

elin

e)

Expected change in Pain Intensity (% relative to baseline)

Expected change in Pain Intensity (% relative to baseline)

Expected change in Pain Intensity (% relative to baseline)

Expected change in Pain Intensity (% relative to baseline)

-10

-5

0

5

10

0 100 200

Cha

nge

in P

45 A

mpl

itude

(µV)

Fig. 3. Expected changes in pain intensity correlated positively with changes in, (a) perceived pain intensity, (b) P45 amplitude, and, (c) P260 amplitude. Expected changes inpain intensity did not correlate with changes in (d) reflex amplitude.

22 P. Goffaux et al. / PAIN� 145 (2009) 18–23

anticipated. Evidence of increased vigilance to nociceptive drives inchronic pain patients has also been reported by Song et al. [28]. Theauthors found that in anticipation of rectal pain, IBS patientsshowed increased fMRI brain activity in the middle frontal gyrusand in SI. This pattern of activity was not observed in healthyadults. Since frontal and posterior parietal regions are closely tiedto sensory associations and vigilance in general, these findings canbe taken to support the presence of heightened attentional pro-cessing when chronic pain is present. We do not yet know if thisform of vigilance develops contemporaneously with the progres-sion of chronic pain or if it predates (and perhaps even predicts)its onset. Future research is necessary to illuminate this possibility.

One of the main implications of this study is that spinal andsupraspinal contributions to anticipated pain relief can be dissoci-ated. Recent research argues that expectations (or placebo effects)are effective because they trigger inhibitory responses that helpdecrease the excitability of spinal neurons [12,15,22,30]. The cur-rent study demonstrates that expectations can produce analgesiaeven in the absence of spinal inhibitory circuits. To our knowledgethis is the first time that the contributions of supraspinal effectorshave been so clearly isolated. Thus, these results support early con-tentions regarding the importance of cognitive re-appraisal duringthe creation of analgesic responses. It is important to point out,however, that these findings do not exclude the possibility ofspinal inhibition during anticipated pain relief, but rather, validatethe idea that supraspinal mechanisms are sufficient to produce theexperience of analgesia.

One potential limitation of this study is worth noting. Duringtesting, the same experimenter recorded patient expectationsand administered the shocks. As a result, possible biases arisingfrom the experimenter’s own expectations may have influencedthe results. Although the use of a strict testing protocol likelycurtailed this effect, future experiments should be aware of this

confound and ensure that testing remains fully blinded. Anotherpotential limitation is the continued use of analgesic/antihyperal-gesic medication by our patients. Although patients in both groupsused similar compounds (and to similar extents), the continueduse of these drugs may have changed their central nervous systemresponses. It is hard to imagine, however, how this potential con-found could explain the pattern of results obtained here (i.e., evi-dence of spinal hyperexcitability in both groups). At best, thesecompounds had no effect on spinal pain signals; at worse, theyproduced central inhibition and prevented the excitability of spinalneurons. Since we only observed facilitatory effects, our results arenot likely attributable to the continued use of analgesic/antihyper-algesic medication. Nevertheless, future studies should includedrug wash-out periods as a prerequisite to testing. Similarly, futurestudies may want to explore why some FM patients naturally ex-pect analgesia, whereas others naturally expect hyperalgesia.Drawing a comprehensive clinical profile for each subgroup willcertainly help to determine why radically different expectationsexist among FM patients. However, despite the importance ofknowing what predicts group membership, predictive factors areunlikely to explain why spinal inhibition was deficient, since bothgroups had impaired descending inhibitory responses.

In conclusion, the findings reported here have important clinicalimplications since they demonstrate that spinal hyperexcitabilitypersists despite subjective pain relief. This means that even whenFM patients consider the immersion procedure as an effectivetreatment, their spine continues to show signs of enhanced pain.This kind of result is never observed when healthy subjects aretested [15]. In fact, when healthy subjects expect and experiencepain relief during the immersion procedure, a significant decreasein their reflex response is observed. Since FM patients showed acompletely different response, this implies a true pathophysiolog-ical change in the processing of their spinal pain signals. A logical

P. Goffaux et al. / PAIN� 145 (2009) 18–23 23

implication of this finding is that FM patients can no longer be sub-sumed under the rubric of somatoform or pseudoneurological dis-orders, a belief still shared by many health care professionals[11,23]. This is because patients suffering from a pseudoneurolog-ical disorder should not demonstrate spinal hyperexcitability,much less when analgesia is expected and reported. On the otherhand, spinal hyperexcitability is entirely possible if a true neuro-logical disorder is present. Even for medical conditions where evi-dence of organicity is scarce, physiological deficits may be present.It is dangerous, therefore, to dismiss reports of pain as psychogenicwhen seemingly thorough investigations fail to demonstrate obser-vable evidence of pathology. Although we caution against general-izing from this small sample, our research, including robust effectsizes, provides objective evidence that FM is essentially a centralnervous system disorder, and not merely a psychological manifes-tation. This constitutes an important and emerging shift in diag-nostic thinking, and one which advocates for neuroprotectivetreatment approaches (also see Kuchinad et al. [19]). Ultimately,it is our hope that such results will promote a greater appreciationof the functional changes underlying chronic widespread pain, thusdecreasing the level of frustration felt by patients and physiciansalike.

Conflict of Interest Statement

The authors declare no competing and financial interests.

Acknowledgements

We thank M.F. Spooner for help in testing and data processing.This work was supported by grants from the Louise EdwardsAwards and the CIHR Placebo Net (S.M. and P.G.). P.G. holds a post-doctoral scholarship from the Fonds de la Recherche en Santé duQuébec, and S.P. holds a postdoctoral scholarship from the Cana-dian Institute of Health Research.

References

[1] Bromm B, Lorenz J. Neurophysiological evaluation of pain. ElectroencephalogrClin Neurophysiol 1998;107:227–53.

[2] Bruner JS, Postman L, Rodrigues J. Expectation and the perception of color. Am JPsychol 1951;64:216–27.

[3] Coffin B, Bouhassira D, Sabate JM, Barbe L, Jian R. Alteration of the spinalmodulation of nociceptive processing in patients with irritable bowelsyndrome. Gut 2004;53:1465–70.

[4] Dowman R. SEP topographies elicited by innocuous and noxious sural nervestimulation. I. Identification of stable periods and individual differences.Electroencephalogr Clin Neurophysiol 1994;92:291–302.

[5] Dowman R. SEP topographies elicited by innocuous and noxious sural nervestimulation. II. Effects of stimulus intensity on topographic pattern andamplitude. Electroencephalogr Clin Neurophysiol 1994;92:303–15.

[6] Dowman R. Effects of interstimulus interval on scalp topographies evoked bynoxious sural nerve stimulation. Psychophysiology 1996;33:398–408.

[7] Dowman R. Pain-evoked anterior cingulate activity generating the negativedifference potential may reflect response selection processes.Psychophysiology 2002;39:369–79.

[8] Dowman R, Darcey TM. SEP topographies elicited by innocuous and noxioussural nerve stimulation. III. Dipole source localization analysis.Electroencephalogr Clin Neurophysiol 1994;92:373–91.

[9] Dowman R, Schell S. The pain-related negative difference potential: a directmeasure of central pain pathway activity or of interactions between theinnocuous somatosensory and pain pathways? Neurophysiol Clin1999;29:423–42.

[10] Dunckley P, Aziz Q, Wise RG, Brooks J, Tracey I, Chang L. Attentionalmodulation of visceral and somatic pain. Neurogastroenterol Motil2007;19:569–77.

[11] Ehrlich GE. Fibromyalgia, a virtual disease. Clin Rheumatol 2003;22:8–11.[12] Fairhurst M, Wiech K, Dunckley P, Tracey I. Anticipatory brainstem activity

predicts neural processing of pain in humans. Pain 2007;128(1–2):101–10.[13] Fields HL, Basbaum AI. Central nervous system mechanisms of pain modu-

lation. In: Wall PD, Melzack R, editors. Textbook of pain. Edinburgh: Churchill;1999. p. 309–29.

[14] Garcia-Larrea L, Frot M, Valeriani M. Brain generators of laser-evokedpotentials: from dipoles to functional significance. Neurophysiol Clin2003;33:279–92.

[15] Goffaux P, redmond WJ, Rainville P, Marchand S. Descending analgesia – whenthe spine echoes what the brain expects. Pain 2007;130:137–43.

[16] Julien N, Goffaux P, Arsenault P, Marchand S. Widespread pain in fibromyalgiais related to a deficit of endogenous pain inhibition. Pain 2005;114:295–302.

[17] Koyama T, Mchaffie JG, Laurienti PJ, Coghill RC. The subjective experience ofpain: where expectations become reality. Proc Natl Acad Sci USA2005;102:12950–5.

[18] Ku Y, Ohara S, Wang L, Lenz FA, Hsiao SS, Bodner M, Hong B, Zhou YD.Prefrontal cortex and somatosensory cortex in tactile crossmodal association:an independent component analysis of ERP recordings. PLoS ONE 2007;2:e771.

[19] Kuchinad A, Schweinhardt P, Seminowicz DA, Wood PB, Chizh BA, BushnellMC. Accelerated brain gray matter loss in fibromyalgia patients: prematureaging of the brain? J Neurosci 2007;27:4004–7.

[20] Lautenbacher S, Rollman GB. Possible deficiencies of pain modulation infibromyalgia. Clin J Pain 1997;13:189–96.

[21] Le Bars D, Dickenson AH, Besson JM. Diffuse noxious inhibitory controls(DNIC). 1. Effects on dorsal horn convergent neurones in the rat. Pain1979;6:283–304.

[22] Matre D, Casey KL, Knardahl S. Placebo-induced changes in spinal cord painprocessing. J Neurosci 2006;26:559–63.

[23] McBeth J, Macfarlane GJ, Benjamin S, Silman AJ. Features of somatizationpredict the onset of chronic widespread pain: results of a large population-based study. Arthritis Rheum 2001;44:940–6.

[24] Paus T. Primate anterior cingulate cortex: where motor control, drive andcognition interface. Nat Rev Neurosci 2001;2:417–24.

[25] Ploghaus A, Tracey I, Gati JS, Clare S, Menon RS, Matthews PM, Rawlins JN.Dissociating pain from its anticipation in the human brain. Science1999;284:1979–81.

[26] Price DD, Barrell JJ. Mechanisms of analgesia produced by hypnosis andplacebo suggestions. Prog Brain Res 2000;122:255–71.

[27] Rainville P. Brain mechanisms of pain affect and pain modulation. Curr OpinNeurobiol 2002;12:195–204.

[28] Song GH, Venkatraman V, Ho KY, Chee MW, Yeoh KG, Wilder-Smith CH.Cortical effects of anticipation and endogenous modulation of visceral painassessed by functional brain MRI in irritable bowel syndrome patients andhealthy controls. Pain 2006;126:79–90.

[29] Tousignant-Laflamme Y, Page S, Goffaux P, Marchand S. An experimentalmodel to measure excitatory and inhibitory pain mechanisms in humans.Brain Res 2008;1230:73–9.

[30] Wager TD, Rilling JK, Smith EE, Sokolik A, Casey KL, Davidson RJ, Kosslyn SM,Rose RM, Cohen JD. Placebo-induced changes in FMRI in the anticipation andexperience of pain. Science 2004;303:1162–7.

[31] Willer JC. Comparative study of perceived pain and nociceptive flexion reflexin man. Pain 1977;3:69–80.

[32] Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C, GoldenbergDI, Tugwell P, Campbell SM, Abeles M, Clark P, Fam AG, Farber SJ,Fiechtner JJ, Franklin CM, Gatter RA, Hamaty D, Lessard J, Lichtbroun AS,Masi AT, Mccain GA, Reynolds WJ, Romano TJ, Russell IJ, Sheon RP. TheAmerian College of Rhumatology 1990 criteria for the classification offibromylagia: report of the Multicenter Criteria Commitee. Arthritis Rheum1990;33:160–72.

[33] Wood PB, Patterson JC, Sunderland JJ, Tainter KH, Glabus MF, Lilien DL.Reduced presynaptic dopamine activity in fibromyalgia syndromedemonstrated with positron emission tomography: a pilot study. J Pain2007;8:51–8.