- Email: [email protected]
The pain inhibiting pain effect: an electrophysiological study in humans
Brain Research 862 (2000) 103–110 www.elsevier.com / locate / bres
Research report
The pain inhibiting pain effect: an electrophysiological study in humans Andreas Reinert, Rolf-Detlef Treede*, Burkhart Bromm Institute of Physiology, University Hospital Eppendorf, Hamburg, Germany Accepted 25 January 2000
Abstract This study examines the counterirritation phenomenon of experimental pain in human subjects. Phasic pain induced by intracutaneous electrical stimuli was simultaneously applied with tonic pain induced by ischemic muscle work. Pain ratings, spontaneous EEG and evoked potentials were measured. We found a significant reduction of phasic pain ratings during and 10 min after tonic pain. The late somatosensory evoked potentials as neurophysiological correlates of phasic pain sensation were attenuated until 20 min after tonic pain offset. The extent of phasic pain relief due to concomitant tonic pain was small but significant, comparable to the effect of a regular systemic dose of a narco-analgesic drug in this experimental pain model. On the other hand, no modulations in the late components of the auditory evoked potential and the power spectrum of the spontaneous EEG were observed. These variables reflect the attention and vigilance of the subject and are well-known to be affected by opioids. The only exception was an increase of beta power, which might reflect hyperarousal during tonic pain. These results support the suggestion, that the analgesic effect of heterotopic noxious stimulation in humans is based on the activation of a specific inhibitory pain control system. Systemic release of endogenous opioids is unlikely to be involved, because the typical effects of opioids on the EEG were not observed. 2000 Elsevier Science B.V. All rights reserved. Themes: Sensory systems Topics: Pain modulation: anatomy and physiology Keywords: Ischemic pain; Intracutaneous stimulus; Somatosensory evoked potentials; Auditory evoked potentials; Spontaneous EEG; Diffuse noxious inhibitory controls
1. Introduction The ‘pain inhibiting pain’ effect is a well-known experience in the folk medicine of many cultures. Meanwhile, several studies have reproduced this counterirritation phenomenon in the human physiological laboratory. Ischemic pain in the arm, for example, elevates the threshold for heat pain on the forehead and the threshold for painful stimulation of the tooth pulp [27]. Many stressful stimuli are able to produce this counterirritation,
*Corresponding author. Present address: Institute of Physiology and Pathophysiology, Johannes Gutenberg University Mainz, Saarstr. 21, D-55099 Mainz, Germany. Tel.: 149-6131-39-25715; fax: 149-6131-3925902. E-mail address: [email protected] (R.-D. Treede)
leading to the term ‘stress induced analgesia’ (SIA). Studies of SIA in humans have used ischemic pain [14,26,27], cold pressor pain [2,33,38], noxious heat [15,36], even long distance running [17] as conditioning stimuli. Significant changes of phasic pain perception have been found by psychophysical techniques [14,27,33,38], human reflex studies [15,26,36] and pain related evoked cerebral potentials [2,14,30,35]. The aim of the present paper was to differentiate between a decrease in specific nociceptive activity and a change in the vigilance level of the subject due to tonic pain. We therefore repeated the measurements described by Chen et al. [14] and examined two more variables, which reflect the momentary attention and vigilance of a person: the late components of the auditory evoked potentials and the power spectrum of the spontaneous EEG [22,29]. The EEG was also analyzed for any changes that might be
0006-8993 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 00 )02077-1
104
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
caused by a strong tonic pain. In addition, we looked for after-effects, which could be another indicator of specific long-lasting changes in the human pain processing systems due to strong tonic pain. For this purpose, the recordings were continued after the end of the ischemia, and a control-session was added to the protocol to eliminate confounding effects of habituation.
intensities were selected (23 pain threshold, 33 pain threshold) and were retained over the whole study. The pain induced by this procedure is described as stabbing, hot and sharp sensation, which differs essentially from the paraesthesiae due to conventionally applied electrical skin shocks [8,32].
2.4. Acoustic stimulus 2. Materials and methods
2.1. Subjects Sixteen healthy male volunteers between 20 and 30 years of age participated in this study. They were carefully briefed and gave informed consent according to the Declaration of Helsinki. The study was approved by the local ethics committee. For each subject the study consisted of three experimental sessions: (1) the adaptation session, the data of which were not evaluated; (2) the control session without tonic pain; (3) the main session with tonic pain.
2.2. Tonic pain stimulus The ischemic pain produced by the submaximum effort tourniquet technique was used as a model for tonic pain (20-min duration). The subject raised his right arm to drain the venous blood. A blood pressure cuff was wrapped around the upper arm and inflated to 200 mmHg. This low pressure was chosen to minimize nerve compression as another source of pain and paraesthesiae [28]. Then, the subject had to perform a defined work by squeezing a hand dynamometer ball to 50–60 mmHg 30 times in 1 min. This muscle work leads to the steadily increasing severe ischemic pain in the lower arm [14]. The cuff was immediately deflated if the ischemic pain became intolerable. This happened in one subject during the adaptation session. This subject therefore did not participate in the study.
2.3. Phasic pain stimulus Brief (20-ms) intracutaneous electrical stimuli were applied to the left middle finger tip to induce phasic test pain in the limb contralateral to the tonic pain. Based on a standardized method, the most superficial layer of a small skin area was removed and a blunt metal electrode was inserted (1 mm in diameter). A large return electrode was wrapped around the finger. Each stimulus consisted of four constant current pulses of 2.5-ms duration. Based on the individual pain threshold determined in the adaptation session, for each subject two clearly painful stimulus-
Three seconds after each intracutaneous stimulus a short acoustic stimulus (20 ms, 500 Hz) prompted the subject to rate the induced pain. Two acoustic stimulus intensities were used, which were cross-modality-matched with the two intracutaneous stimulus intensities in the adaptation session. Thus, for every subject the acoustic stimulus intensities were as loud as the intracutaneous stimulus intensities were painful and the corresponding late evoked potential amplitudes were similar across the two modalities (see Results).
2.5. Pain rating scale Sensation was rated on the same open-ended scale for tonic ischemia and for phasic intracutaneous stimuli: 05no sensation; 15faint sensation, non-painful; 25mild sensation, non-painful; 35strong sensation, non-painful; 45 faint pain; 55mild pain; 65strong pain; 75very strong pain; 8 and more5extreme but still tolerable pain.
2.6. General experimental protocol In the adaptation session, the subjects were accustomed to all stimuli and the whole set-up. In the same session, the individual pain threshold was determined and the acoustic stimulus intensities were chosen with the described crossmodality-match. The two main experiments are schematically shown in Fig. 1. The ischemia experiment consisted of six stimulus blocks of 10 min duration each. In each block, 40 intracutaneous and 40 acoustic stimuli were given. The lower part of the figure shows a segment of 5 min as an example. The intensity of intracutaneous and acoustic stimuli was independently randomized and also the interstimulus intervals were randomized to minimize habituation. Two blocks were recorded before the ischemic pain, two blocks during ischemic pain and two blocks after ischemic pain. In this and all the following figures the horizontal black bar shows the ischemia duration. Note that the ischemia was produced in the arm contralateral to the finger stimulated electrically. The design of the control experiment corresponds exactly to the ischemia experiment, even the defined work was done, only the cuff was deflated before this work.
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
105
For the frequency analysis of the spontaneous EEG, segments of 2.56-s duration immediately before each intracutaneous stimulus were digitized at 100 Hz. These segments were subjected to a Fast-Fourier-Transform (Kaiser window 240 dB), and the integrated power was computed in the classical frequency ranges: delta 1–3.9 Hz; theta 4–7.4 Hz; alpha 1 7.5–8.9 Hz; alpha 2 9–12.4 Hz; beta 12.5–18 Hz. For statistical analyses, a two-way analysis of variance for repeated measures was performed, with the factors ‘time course’ (the six stimulus blocks) and ‘experimental condition’ (control, ischemic pain). Differences between the control and ischemia experiment were then located by two-tailed paired t-tests.
3. Results
3.1. Models for tonic and phasic pain
Fig. 1. Experimental protocol. The upper part of the figure shows the differences between control and ischemia experiment. In the control experiment the cuff was deflated before the defined work was done (hatched area). The black bar shows the duration of the ischemia in the ischemia experiment (20 min). The middle part of this figure demonstrates the division of both experimental sessions into six stimulus blocks of 10-min duration each. Two stimulus blocks were recorded before the ischemic pain, two blocks during pain and two blocks in the post-ischemic period. In each stimulus block, 40 intracutaneous and 40 acoustic stimuli were given in randomized order. The lower part of the figure shows a segment of 5 min as an example.
Fig. 2A shows the mean ratings of the tonic pain (n516). After the defined work, pain began to rise and after 20 min of ischemia all subjects experienced a very strong pain. When the same work was performed without ischemia in the control session, it was not painful. Likewise, ischemia without muscle work is not painful. The ischemic muscle pain was immediately relieved by deflation of the cuff. A few seconds later, however, postischemic paraesthesiae appeared with a maximum after 2 min which were slightly painful in most subjects (n514). The tingling and pricking paraesthesiae remained detectable for 762 min (mean6S.D.). The pain threshold for the intracutaneous stimulus in the control session was 0.3860.19 mA and in the ischemia session 0.3960.18 mA (correlation coefficient r50.76).
2.7. EEG recording and data analysis The EEG was recorded continuously during the whole experiments over the vertex (Cz) versus linked ear lobes (bandpass 1–30 Hz). The data were stored on magnetic tape and were digitized off-line (100 Hz). For every stimulus block, stimulus synchronous averages were computed and plotted. For artefact control, the EOG was recorded between supra- and infraorbital electrodes using the same bandpass. EEG segments with EOG contamination were automatically excluded from all analyses. To verify the absence of contamination by eye movements, the EOG was also averaged for every stimulus block and plotted. The peaks of the late components of the evoked potentials were determined, and the peak-to-peak amplitude was measured: The SEP showed a negativity at about 150 ms and a positivity at about 250 ms, the AEP had its negativity at 100 ms and its positivity at about 200 ms.
3.2. Phasic pain during tonic pain Phasic pain ratings in response to the intracutaneous stimuli (Fig. 2B) were generally less than those of tonic pain. During the second half of the ischemia, they were significantly reduced by 17% in comparison with the control experiment (open symbols). The formerly clearly painful intracutaneous stimulus now elicited only a touch sensation, which means even a reduction by 100% with respect to the painfulness of the stimulus. Shortly after the deflation of the cuff, the intracutaneous stimulus became painful again, but the ratings were still significantly reduced by 9%. In the sixth block the phasic pain ratings showed a complete recovery. In the control experiment one can see a slight reduction of the pain ratings over the 60-min of the experiment, presumably due to habituation. Because of the limited
106
A. Reinert et al. / Brain Research 862 (2000) 103 – 110 Table 1 Analysis of variance Variable
Rating SEP a AEP a Delta Theta Alpha 1 Alpha 2 Beta
F Block 1–6 (S)
Contr.–isch. (R)
Interaction (RS)
38.31*** 48.38*** 23.36*** 2.11 1.50 0.56 1.35 25.25***
0.74 9.83** 2.64 1.96 0.98 0.29 0.70 9.81**
9.36*** 4.69*** 1.73 1.42 2.28 0.28 0.58 6.89***
a
SEP: somatosensory evoked potential AEP: auditory evoked potential. *** P#0.001; ** P#0.01; * P#0.05.
(habituation) and a significant interaction (Table 1). The significant interaction indicates, that at some time points there was a significant difference between the control and ischemia experiment (see above).
3.3. Somatosensory and auditory evoked potentials during tonic pain
Fig. 2. (A) Time-course of the ischemic pain. Mean pain ratings and their standard deviations (n516) during (circles) and immediately after (squares) a 10-min ischemia period (horizontal filled bar). The subjects had to rate the pain every 2.5 min (rating ,4: non painful, rating $4: painful). The deflation of the cuff was followed by immediate relief of pain, which was replaced by post-ischemic paraesthesiae. (B) Phasic pain ratings before, during and after tonic pain (filled symbols). The horizontal dashed line divides non-painful (rating ,4) and painful sensations (rating $4). Note the significant reduction of phasic pain ratings compared to the control session (open symbols). (C) Late somatosensory evoked potentials (peak-to-peak amplitudes). Note the reduction of the SEP-amplitudes during the ischemic pain and the long lasting after-effect. (D) Late auditory evoked potentials (peak-to-peak amplitudes). Note, that there is no significant difference between the AEP-amplitudes from the control and ischemia experiment. Mean values and standard deviations (n516) from control (open symbols) and ischemia experiment (filled symbols). Every symbol stands for one stimulus block. The filled horizontal bar indicates the ischemia duration in the ischemia experiment. Significant differences between the two experimental conditions (two tailed t-test) are marked by asterisks: *** P#0.001; ** P#0.01; * P#0.05.
The effects on the neurophysiological level of measurement were much more distinct (Fig. 2C and Table 1). Like the subjective pain ratings, the SEP-amplitudes showed a habituation across the course of both the control (open symbols) and the ischemia session (filled symbols). But already in the first block during the ischemia there was a significant reduction of these amplitudes by 28% compared with the control experiment. This effect remained in the second half of the ischemia with a reduction by 27% and became still larger after the end of the ischemia period with a reduction by 35%. In the last stimulus block there was a clear increase of the SEP-amplitudes in the ischemia experiment, but the difference between the two experimental conditions remained significant. This long-lasting aftereffect (20 min) of the ischemia shown in the pain related components of the SEP is remarkable. The two-way ANOVA for the AEP showed no effect of tonic pain, only a significant effect of the time course (Table 1). Fig. 2D illustrates that this is again due to a slight habituation, which was similar in the control experiment and under severe tonic pain. Throughout all six stimulus blocks, AEP amplitudes in control and ischemia session were similar. Table 2 Relative changes (post-pre) / pre controlischemia a
ischemia time and the attempt to show a possible aftereffect, the intracutaneous stimuli had to be applied continuously without the break needed for dishabituation. To account for habituation, all statistical tests were performed versus the corresponding time point of the control experiment. Statistical analysis was performed by two-way ANOVA showing a significant effect of the time course
E SEP AEP bEEG a
Control (%)
Ischemia (%)
25.5 213.7 210.0 24.4
226.1 230.6 217.0 148.8
All comparisons were made between block 2 (immediately before ischemia) and block 4 (during maximal ischemic pain).
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
107
Fig. 3. Changes of EEG power during tonic pain. Mean power values and standard deviations (n516) in the classical frequency bands from the control (open bars) and ischemia experiment (filled bars). Every bar stands for one stimulus block. The filled horizontal bar indicates the ischemia duration in the ischemia experiment. (A) Delta (1–3.9 Hz); (B) theta (4–7.4 Hz); (C) alpha 1 (7.5–8.9 Hz); (D) alpha 2 (9–12.4 Hz); (E) beta (12.5–18 Hz). Note the significant increase of beta power during and shortly after ischemic pain. Statistically significant differences between the two experimental conditions (two tailed t-tests) are marked by asterisks: *** P#0.001; ** P#0.01; * P#0.05.
3.4. EEG during tonic pain
4. Discussion
The two-way ANOVA for the power spectrum of the spontaneous EEG showed significant effects only in the beta frequency range. Details are given in Fig. 3: the beta power increased significantly during ischemic pain in comparison with the control experiment. This increase remained during the first ten min after ischemic pain. In the last stimulus block, the beta-power again decreased to control levels. Increases in delta or theta power, which would reflect reduced vigilance, were not observed in this study. To the contrary, there was a small but non-significant decrease around the end of ischemic pain (stimulus blocks 4 and 5). Power in the alpha frequency bands remained remarkably stable throughout the study. Notably, there was also no trend for a shift of power from fast alpha to slow alpha.
The present study examined the counterirritation phenomenon between two painful stimuli using phasic pain, caused by intracutaneous electrical stimuli of one hand, and a concurrent tonic pain caused by ischemia of the other arm. Phasic pain and its neurophysiological correlate, late components of the somatosensory evoked potential, were found to be significantly reduced by heterotopic tonic pain. Pain reduction outlasted the 20-min conditioning stimuli by another 10–20 min. The extent of pain reduction was similar to a regular systemic dose of an opioid drug [32]. On the other hand, neurophysiological control recordings of the auditory evoked potentials and the spontaneous EEG gave no evidence for reduced vigilance or attention during tonic pain. The only EEG effect of the strong tonic pain was a small increase in beta power. The observed after-effect indicates long-lasting changes in the human pain processing systems, triggered by strong tonic pain. After the release of the cuff, there was a sudden relief from ischemic pain, but the subjective ratings of the phasic pain remained reduced for another 10 min. The SEPs showed their largest reduction immediately after the end of the ischemia and remained significantly reduced for 20 min. Similar after-effects of 5–60-min duration were found for the human blink reflex, for the muscular pain threshold [26], for nociceptive reflexes [36] and for heat
3.5. Habituation Table 2 shows the relative changes in all parameters from immediately before ischemia to the time of maximal ischemic pain. The values from the control session document some habituation (around 10%) due to the long lasting sessions with continuous stimulation. But the effects of tonic pain were clearly greater than the effects due to habituation.
108
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
pain ratings [33]. On the other hand, Price and McHaffie [31] reported no after-effect when they reduced the pain sensation of electrical test stimuli by simultaneously applied noxious heat stimuli. Probably, the very short duration (10 s) of their conditioning heat stimuli was responsible for this outcome [15,34,36]. A previous study in our laboratory also failed to document significant aftereffects after 15 min of ischemic pain [14]. But in that study measurements were interrupted for several minutes after each stimulus block to allow for dishabituation, and an after-effect may thus have been missed. In conclusion, with conditioning stimuli of sufficient intensity and duration, long lasting effects on inhibitory pain control systems have regularly been found. The time immediately after release of the cuff is characterized by the appearance of the tingling and pricking post-ischemic paraesthesiae. Microneurographic studies in humans have shown that these paraesthesiae are due to ectopically generated paroxysmal discharges in myelinated fibres; unmyelinated fibres are not involved [25]. This afferent barrage in large myelinated fibres could be involved in the observed after-effects. Pain inhibition by activation of Ab-fibres is the basis of transcutaneous electrical nerve stimulation (classical TENS) and dorsal column stimulation. With these techniques, pain relief is only obtained when the paraesthesiae are projected to the same region as the pain [24,39]. Also the pain relief by vibration is more efficient when applied within the area of pain than when applied to neighbouring areas, contralateral vibration had no effect [5,21]. In our experiments, the post-ischemic paraesthesiae and the phasic test pain were located in contralateral body areas. Thus, local or segmental interactions between large and small diameter fibres can be ruled out as the source of pain inhibition in the postischemia period. The counterirritation phenomenon reported here resembles inhibitory mechanisms in the animal model, termed ‘diffuse noxious inhibitory controls’ (DNIC). Spinal convergent neurons in the rat are inhibited by painful stimulation of body areas outside the receptive field of these neurons [20]. Especially C-fibre stimulation activates these inhibitory control systems, the same fibre population which is excited by ischemic muscle contractions [23]. Long lasting after-effects are characteristic of DNIC. The inhibition is thought to be mediated by a supraspinal loop with a complex pharmacology (for review see Refs. [16,19]). The role of the endogenous opioids in pain control systems is especially well investigated. In animal experiments two different forms of foot shock induced analgesia were found, one opiate mediated and the other without the participation of opiates. Which of both systems will be activated seems to depend on stimulus parameters: The higher the intensity and the longer the duration of the painful stimuli, the more likely it is, that the opiate antagonist naloxone will antagonize the analgesia induced
by exposure to those stimuli [1]. In humans, experiments with naloxone gave contradictory results: Willer et al. [37] demonstrated a naloxone-reversible inhibition of nociceptive withdrawal reflexes during the anticipation of a very strong foot shock. The pain threshold elevation induced by cold pressor pain was naloxone-reversible [18]. The analgesic effect of a long distance run on the perception of ischemic pain was also naloxone-reversible, but in the same experiment, the analgesic effect on noxious thermal stimulation was not [17]. Pertovaara et al. [27] found no effect of naloxone on the increase of the dental pain threshold during ischemic pain. Obviously, also in humans there are different pain control systems which can be activated by noxious stimuli. The effects of tonic pain on the ratings of phasic pain (26% reduction with respect to baseline) met the range of the effects of strong narcotic analgesics. Under similar experimental conditions, tilidine (100 mg, p.o.) decreased phasic pain ratings by 25% in comparison to placebo [9], meperidine (150 mg, p.o.) by 40% [10], and pentazocine (30 mg, i.v.) by 15% [7]. Chapman et al. [12] described decreases in subjective pain report by 23% following fentanyl (0.1 mg, i.v.) and by 25% following nitrous oxide (33% in oxygen). Our control data, however, provide indirect evidence against the involvement of endogenous opioids in the analgesia caused by ischemic pain. Opioids and other centrally acting analgesics affect the spontaneous EEG in a characteristic way: at low doses they slow the dominant alpha frequency, at higher doses they reduce the power in the alpha band and increase the activity in the delta and theta bands [6,11,32]. If endogenous opioids are released during strong tonic pain, we would expect similar changes in the EEG of our subjects. But we found the opposite EEG alterations, a slight decrease of power in the slow frequency range, accompanied by a statistically significant increase of beta power. Similar EEG changes were found during cold pressor pain [3] and in patients suffering from dental pain [13]. In PET studies, ischemic pain did not lead to a displacement of exogenous opioid-receptor ligands (Treede, Campbell, Magiros, and Frost, unpublished data), in contrast to capsaicin-induced pain [4]. These data argue against the action of endogenous opioids at cerebral opioid receptors during ischemic pain. The absence of the typical opioid effects on the EEG during ischemic pain may also indicate that endogenous opioids are only released locally at an appropriate site in the central nervous system, e.g. in the spinal cord dorsal horn. The described EEG alterations gave evidence against a decreased state of vigilance in our subjects during the experiments. Such a decrease of vigilance would lead to a reduction of power in the alpha range and a shift toward lower frequencies [22]. Also the late components of the auditory evoked potentials are strongly influenced by the state of alertness of the subject [29]. We did not observe
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
any significant change in the late components of the AEP during and after strong ischemic pain. In a study on the inhibition of proximal pain perception by distal coldpressor pain, it was argued that the ipsilateral effects were distinct from a change in attention, because they were more pronounced than the contralateral effects [2]. Our data indicate that contralateral pain inhibition is also distinct from distraction. In summary, the inhibition of a phasic test pain by simultaneous tonic pain is not due to altered attention or vigilance, because there were no corresponding changes in AEPs or EEG spectra. The long-lasting after effect (longer than 10 min), which is not explained by segmental mechanisms caused by post-ischemic paraesthesiae, indicates that this inhibition is likely due to the activation of a specific endogenous pain control system. Cerebral release of endogenous opioids, however, would have to be highly localized to this part of the endogenous pain control system, because we found none of the known opioid effects in the spontaneous EEG.
References [1] Z. Amit, Z.H. Galina, Stress-induced analgesia: adaptive pain suppression, Physiol. Rev. 66 (1986) 1091–1120. [2] L. Arendt-Nielsen, K. Gotliebsen, Segmental inhibition of laserevoked brain potentials by ipsi- and contralaterally applied cold pressor pain, Eur. J. Appl. Physiol. 64 (1992) 56–61. [3] M. Backonja, E.W. Howland, J. Wang, J. Smith, M. Salinsky, C.S. Cleeland, Tonic changes in alpha power during immersion of the hand in cold water, Electroenceph. Clin. Neurophysiol. 79 (1991) 192–203. [4] B. Bencherif, M. Oropeza, P.N. Fuchs, R.N. Sheth, S.T. Li, H. Ravert, J. Musachio, B. Mathews, B. Dannals, J.N. Campbell, J.J. Frost, PET scan analysis of m opioid receptor binding during a tonic painful stimulus in normal human subjects, in: Abstracts of 9th World Congress on Pain, Wien, 1999, p. 281. ¨ [5] G. Bini, G. Cruccu, K.-E. Hagbarth, W. Schady, E. Torebjork, Analgesic effect of vibration and cooling on pain induced by intraneural electrical stimulation, Pain 18 (1984) 239–248. [6] T.A. Bowdle, R.J. Ward, Induction of anesthesia with small doses of sufentanil or fentanyl: dose versus EEG response, speed of onset, and thiopental requirement, Anesthesiology 70 (1989) 26–30. [7] B. Bromm, R. Ganzel, W.M. Herrmann, W. Meier, E. Scharein, The analgesic efficacy of flupirtine in comparison to pentazocine and placebo assessed by EEG and subjective pain ratings, Postgrad. Med. J. 63 (1987) 109–112. [8] B. Bromm, W. Meier, The intracutaneous stimulus: a new pain model for algesimetric studies, Methods Find. Exp. Clin. Pharmacol. 6 (1984) 405–410. [9] B. Bromm, W. Meier, E. Scharein, Antagonism between tilidine and naloxone on cerebral potentials and pain ratings in man, Eur. J. Pharmacol. 87 (1983) 431–439. [10] B. Bromm, W. Meier, E. Scharein, Imipramine reduces experimental pain, Pain 25 (1986) 245–257. [11] B. Bromm, W. Meier, E. Scharein, Pre-stimulus / post-stimulus relations in EEG spectra and their modulations by an opioid and an antidepressant, Electroenceph. Clin. Neurophysiol. 73 (1989) 188– 197.
109
[12] C.R. Chapman, Y.M. Colpitts, C. Benedetti, S. Butler, Event-related potential correlates of analgesia; comparison of fentanyl, acupuncture, and nitrous oxide, Pain 14 (1982) 327–337. [13] A.C. Chen, S.F. Dworkin, M.T. Drangsholt, Cortical power spectral analysis of acute pathophysiological pain, Int. J. Neurosci. 18 (1983) 269–278. [14] A.C.N. Chen, R.-D. Treede, B. Bromm, Tonic pain inhibits phasic pain: Evoked cerebral potential correlates in man, Psychiatry Res. 14 (1985) 343–351. [15] J. Ellrich, R.-D. Treede, Characterization of blink reflex interneurons by activation of diffuse noxious inhibitory controls in man, Brain Res. 803 (1998) 161–168. [16] H.L. Fields, A.J. Basbaum, Endogenous pain control mechanisms, in: P.D. Wall, R. Melzack (Eds.), Textbook of Pain, Churchill Livingstone, Edinburgh, 1989, pp. 206–217. [17] M.N. Janal, E.W.D. Colt, W.C. Clark, M. Glusman, Pain sensitivity, mood and plasma endocrine levels in man following long-distance running: effects of naloxone, Pain 19 (1984) 13–25. [18] G. Jungkunz, R.R. Engel, U.G. King, H.J. Kuss, Endogenous opiates increase pain tolerance after stress in humans, Psychiatry Res. 8 (1983) 13–18. [19] D. Le Bars, D. Bouhassira, L. Villanueva, Opioids and diffuse noxious inhibitory control (DNIC) in the rat, in: B. Bromm, J.E. Desmedt (Eds.), Advances in Pain Research and Therapy, Raven Press, New York, 1995, pp. 517–539. [20] D. Le Bars, L. Villanueva, Electrophysiological evidence for the activation of descending inhibitory controls by nociceptive afferent pathways, Prog. Brain Res. 77 (1988) 275–299. ˚ [21] T. Lundeberg, D. Ottoson, S. Hakansson, B.A. Meyerson, Vibratory stimulation for the control of intractable chronic orofacial pain, in: J.J. Bonica, U. Lindblom, A. Iggo (Eds.), Raven Press, New York, 1983, pp. 555–561. [22] M. Matejcek, Vigilance and the EG: psychological, physiological and pharmacological aspects, in: W.M. Herrmann (Ed.), Electroencephalography in Drug Research, Gustav Fischer, Stuttgart, 1982, pp. 405–508. [23] S. Mense, Nociception from skeletal muscle in relation to clinical muscle pain, Pain 54 (1993) 241–289. [24] R.B. North, D.H. Kidd, M. Zahurak, C.S. James, D.M. Long, Spinal cord stimulation for chronic, intractable pain: experience over two decades, Neurosurgery 32 (1993) 384–395. ¨ [25] J.L. Ochoa, H.E. Torebjork, Paraesthesiae from ectopic impulse generation in human sensory nerves, Brain 103 (1980) 835–853. [26] T. Pantaleo, R. Duranti, F. Bellini, Effects of heterotopic ischemic pain on muscular pain threshold and blink relfex in humans, Neurosci. Lett. 85 (1988) 56–60. [27] A. Pertovaara, P. Kemppainen, G. Johansson, S.-L. Karonen, Ischemic pain nonsegmentally produces a predominant reduction of pain and thermal sensitivity in man: a selective role for endogenous opioids, Brain Res. 251 (1982) 83–92. ¨ [28] A. Pertovaara, T. Nurmikko, P.J. Pontinen, Two separate components of pain produced by the submaximal effort tourniquet test, Pain 20 (1984) 53–58. [29] T.W. Picton, S.A. Hillyard, R. Galambos, Habituation and attention in the auditory system, in: W.D. Keidel, W.D. Neff (Eds.), Handbook of Sensory Physiology. Auditory System. Clinical and Special Topics, Springer, Berlin, 1976, pp. 343–389. [30] L. Plaghki, D. Delisle, J.M. Godfraind, Heterotopic nociceptive conditioning stimuli and mental task modulate differently the perception and physiological correlates of short CO 2 laser stimuli, Pain 57 (1994) 181–192. [31] D.D. Price, J.G. McHaffie, Effects of heterotopic conditioning stimuli on first and second pain: a psychophysical evaluation in humans, Pain 34 (1988) 245–252. [32] E. Scharein, B. Bromm, The intracutaneous pain model in the assessment of analgesic efficacy, Pain Rev. 5 (1998) 216–246.
110
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
[33] J.D. Talbot, G.H. Duncan, M.C. Bushnell, M. Boyer, Diffuse noxious inhibitory controls (DNICs): psychophysical evidence in man for intersegmental suppression of noxious heat perception by cold pressor pain, Pain 30 (1987) 221–232. [34] L. Villanueva, D. Le Bars, The encoding of thermal stimuli applied to the tail of the rat by lowering the excitability of trigeminal convergent neurones, Brain Res. 330 (1985) 245–251. [35] S. Watanabe, R. Kakigi, M. Hoshiyama, Y. Kitamura, S. Koyama, M. Shimojo, Effects of noxious cooling of the skin on pain perception in man, J. Neurol. Sci. 135 (1996) 68–73. [36] J.C. Willer, T. De Broucker, D. Le Bars, Encoding of nociceptive
thermal stimuli by diffuse noxious inhibitory controls in humans, J. Neurophysiol. 62 (1989) 1028–1038. [37] J.C. Willer, H. Dehen, J. Cambier, Stress-induced analgesia in humans: endogenous opioids and naloxone-reversible depression of pain reflexes, Science 212 (1981) 689–691. [38] N. Witting, P. Svensson, L. Arendt-Nielsen, T.S. Jensen, Differential effect of painful heterotopic stimulation on capsaicin-induced pain and allodynia, Brain Res. 801 (1998) 206–210. [39] C.J. Woolf, Transcutaneous and implanted nerve stimulation, in: P.D. Wall, R. Melzack (Eds.), Textbook of Pain, Churchill Livingstone, Edinburgh, 1984, pp. 679–690.
Research report
The pain inhibiting pain effect: an electrophysiological study in humans Andreas Reinert, Rolf-Detlef Treede*, Burkhart Bromm Institute of Physiology, University Hospital Eppendorf, Hamburg, Germany Accepted 25 January 2000
Abstract This study examines the counterirritation phenomenon of experimental pain in human subjects. Phasic pain induced by intracutaneous electrical stimuli was simultaneously applied with tonic pain induced by ischemic muscle work. Pain ratings, spontaneous EEG and evoked potentials were measured. We found a significant reduction of phasic pain ratings during and 10 min after tonic pain. The late somatosensory evoked potentials as neurophysiological correlates of phasic pain sensation were attenuated until 20 min after tonic pain offset. The extent of phasic pain relief due to concomitant tonic pain was small but significant, comparable to the effect of a regular systemic dose of a narco-analgesic drug in this experimental pain model. On the other hand, no modulations in the late components of the auditory evoked potential and the power spectrum of the spontaneous EEG were observed. These variables reflect the attention and vigilance of the subject and are well-known to be affected by opioids. The only exception was an increase of beta power, which might reflect hyperarousal during tonic pain. These results support the suggestion, that the analgesic effect of heterotopic noxious stimulation in humans is based on the activation of a specific inhibitory pain control system. Systemic release of endogenous opioids is unlikely to be involved, because the typical effects of opioids on the EEG were not observed. 2000 Elsevier Science B.V. All rights reserved. Themes: Sensory systems Topics: Pain modulation: anatomy and physiology Keywords: Ischemic pain; Intracutaneous stimulus; Somatosensory evoked potentials; Auditory evoked potentials; Spontaneous EEG; Diffuse noxious inhibitory controls
1. Introduction The ‘pain inhibiting pain’ effect is a well-known experience in the folk medicine of many cultures. Meanwhile, several studies have reproduced this counterirritation phenomenon in the human physiological laboratory. Ischemic pain in the arm, for example, elevates the threshold for heat pain on the forehead and the threshold for painful stimulation of the tooth pulp [27]. Many stressful stimuli are able to produce this counterirritation,
*Corresponding author. Present address: Institute of Physiology and Pathophysiology, Johannes Gutenberg University Mainz, Saarstr. 21, D-55099 Mainz, Germany. Tel.: 149-6131-39-25715; fax: 149-6131-3925902. E-mail address: [email protected] (R.-D. Treede)
leading to the term ‘stress induced analgesia’ (SIA). Studies of SIA in humans have used ischemic pain [14,26,27], cold pressor pain [2,33,38], noxious heat [15,36], even long distance running [17] as conditioning stimuli. Significant changes of phasic pain perception have been found by psychophysical techniques [14,27,33,38], human reflex studies [15,26,36] and pain related evoked cerebral potentials [2,14,30,35]. The aim of the present paper was to differentiate between a decrease in specific nociceptive activity and a change in the vigilance level of the subject due to tonic pain. We therefore repeated the measurements described by Chen et al. [14] and examined two more variables, which reflect the momentary attention and vigilance of a person: the late components of the auditory evoked potentials and the power spectrum of the spontaneous EEG [22,29]. The EEG was also analyzed for any changes that might be
0006-8993 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 00 )02077-1
104
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
caused by a strong tonic pain. In addition, we looked for after-effects, which could be another indicator of specific long-lasting changes in the human pain processing systems due to strong tonic pain. For this purpose, the recordings were continued after the end of the ischemia, and a control-session was added to the protocol to eliminate confounding effects of habituation.
intensities were selected (23 pain threshold, 33 pain threshold) and were retained over the whole study. The pain induced by this procedure is described as stabbing, hot and sharp sensation, which differs essentially from the paraesthesiae due to conventionally applied electrical skin shocks [8,32].
2.4. Acoustic stimulus 2. Materials and methods
2.1. Subjects Sixteen healthy male volunteers between 20 and 30 years of age participated in this study. They were carefully briefed and gave informed consent according to the Declaration of Helsinki. The study was approved by the local ethics committee. For each subject the study consisted of three experimental sessions: (1) the adaptation session, the data of which were not evaluated; (2) the control session without tonic pain; (3) the main session with tonic pain.
2.2. Tonic pain stimulus The ischemic pain produced by the submaximum effort tourniquet technique was used as a model for tonic pain (20-min duration). The subject raised his right arm to drain the venous blood. A blood pressure cuff was wrapped around the upper arm and inflated to 200 mmHg. This low pressure was chosen to minimize nerve compression as another source of pain and paraesthesiae [28]. Then, the subject had to perform a defined work by squeezing a hand dynamometer ball to 50–60 mmHg 30 times in 1 min. This muscle work leads to the steadily increasing severe ischemic pain in the lower arm [14]. The cuff was immediately deflated if the ischemic pain became intolerable. This happened in one subject during the adaptation session. This subject therefore did not participate in the study.
2.3. Phasic pain stimulus Brief (20-ms) intracutaneous electrical stimuli were applied to the left middle finger tip to induce phasic test pain in the limb contralateral to the tonic pain. Based on a standardized method, the most superficial layer of a small skin area was removed and a blunt metal electrode was inserted (1 mm in diameter). A large return electrode was wrapped around the finger. Each stimulus consisted of four constant current pulses of 2.5-ms duration. Based on the individual pain threshold determined in the adaptation session, for each subject two clearly painful stimulus-
Three seconds after each intracutaneous stimulus a short acoustic stimulus (20 ms, 500 Hz) prompted the subject to rate the induced pain. Two acoustic stimulus intensities were used, which were cross-modality-matched with the two intracutaneous stimulus intensities in the adaptation session. Thus, for every subject the acoustic stimulus intensities were as loud as the intracutaneous stimulus intensities were painful and the corresponding late evoked potential amplitudes were similar across the two modalities (see Results).
2.5. Pain rating scale Sensation was rated on the same open-ended scale for tonic ischemia and for phasic intracutaneous stimuli: 05no sensation; 15faint sensation, non-painful; 25mild sensation, non-painful; 35strong sensation, non-painful; 45 faint pain; 55mild pain; 65strong pain; 75very strong pain; 8 and more5extreme but still tolerable pain.
2.6. General experimental protocol In the adaptation session, the subjects were accustomed to all stimuli and the whole set-up. In the same session, the individual pain threshold was determined and the acoustic stimulus intensities were chosen with the described crossmodality-match. The two main experiments are schematically shown in Fig. 1. The ischemia experiment consisted of six stimulus blocks of 10 min duration each. In each block, 40 intracutaneous and 40 acoustic stimuli were given. The lower part of the figure shows a segment of 5 min as an example. The intensity of intracutaneous and acoustic stimuli was independently randomized and also the interstimulus intervals were randomized to minimize habituation. Two blocks were recorded before the ischemic pain, two blocks during ischemic pain and two blocks after ischemic pain. In this and all the following figures the horizontal black bar shows the ischemia duration. Note that the ischemia was produced in the arm contralateral to the finger stimulated electrically. The design of the control experiment corresponds exactly to the ischemia experiment, even the defined work was done, only the cuff was deflated before this work.
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
105
For the frequency analysis of the spontaneous EEG, segments of 2.56-s duration immediately before each intracutaneous stimulus were digitized at 100 Hz. These segments were subjected to a Fast-Fourier-Transform (Kaiser window 240 dB), and the integrated power was computed in the classical frequency ranges: delta 1–3.9 Hz; theta 4–7.4 Hz; alpha 1 7.5–8.9 Hz; alpha 2 9–12.4 Hz; beta 12.5–18 Hz. For statistical analyses, a two-way analysis of variance for repeated measures was performed, with the factors ‘time course’ (the six stimulus blocks) and ‘experimental condition’ (control, ischemic pain). Differences between the control and ischemia experiment were then located by two-tailed paired t-tests.
3. Results
3.1. Models for tonic and phasic pain
Fig. 1. Experimental protocol. The upper part of the figure shows the differences between control and ischemia experiment. In the control experiment the cuff was deflated before the defined work was done (hatched area). The black bar shows the duration of the ischemia in the ischemia experiment (20 min). The middle part of this figure demonstrates the division of both experimental sessions into six stimulus blocks of 10-min duration each. Two stimulus blocks were recorded before the ischemic pain, two blocks during pain and two blocks in the post-ischemic period. In each stimulus block, 40 intracutaneous and 40 acoustic stimuli were given in randomized order. The lower part of the figure shows a segment of 5 min as an example.
Fig. 2A shows the mean ratings of the tonic pain (n516). After the defined work, pain began to rise and after 20 min of ischemia all subjects experienced a very strong pain. When the same work was performed without ischemia in the control session, it was not painful. Likewise, ischemia without muscle work is not painful. The ischemic muscle pain was immediately relieved by deflation of the cuff. A few seconds later, however, postischemic paraesthesiae appeared with a maximum after 2 min which were slightly painful in most subjects (n514). The tingling and pricking paraesthesiae remained detectable for 762 min (mean6S.D.). The pain threshold for the intracutaneous stimulus in the control session was 0.3860.19 mA and in the ischemia session 0.3960.18 mA (correlation coefficient r50.76).
2.7. EEG recording and data analysis The EEG was recorded continuously during the whole experiments over the vertex (Cz) versus linked ear lobes (bandpass 1–30 Hz). The data were stored on magnetic tape and were digitized off-line (100 Hz). For every stimulus block, stimulus synchronous averages were computed and plotted. For artefact control, the EOG was recorded between supra- and infraorbital electrodes using the same bandpass. EEG segments with EOG contamination were automatically excluded from all analyses. To verify the absence of contamination by eye movements, the EOG was also averaged for every stimulus block and plotted. The peaks of the late components of the evoked potentials were determined, and the peak-to-peak amplitude was measured: The SEP showed a negativity at about 150 ms and a positivity at about 250 ms, the AEP had its negativity at 100 ms and its positivity at about 200 ms.
3.2. Phasic pain during tonic pain Phasic pain ratings in response to the intracutaneous stimuli (Fig. 2B) were generally less than those of tonic pain. During the second half of the ischemia, they were significantly reduced by 17% in comparison with the control experiment (open symbols). The formerly clearly painful intracutaneous stimulus now elicited only a touch sensation, which means even a reduction by 100% with respect to the painfulness of the stimulus. Shortly after the deflation of the cuff, the intracutaneous stimulus became painful again, but the ratings were still significantly reduced by 9%. In the sixth block the phasic pain ratings showed a complete recovery. In the control experiment one can see a slight reduction of the pain ratings over the 60-min of the experiment, presumably due to habituation. Because of the limited
106
A. Reinert et al. / Brain Research 862 (2000) 103 – 110 Table 1 Analysis of variance Variable
Rating SEP a AEP a Delta Theta Alpha 1 Alpha 2 Beta
F Block 1–6 (S)
Contr.–isch. (R)
Interaction (RS)
38.31*** 48.38*** 23.36*** 2.11 1.50 0.56 1.35 25.25***
0.74 9.83** 2.64 1.96 0.98 0.29 0.70 9.81**
9.36*** 4.69*** 1.73 1.42 2.28 0.28 0.58 6.89***
a
SEP: somatosensory evoked potential AEP: auditory evoked potential. *** P#0.001; ** P#0.01; * P#0.05.
(habituation) and a significant interaction (Table 1). The significant interaction indicates, that at some time points there was a significant difference between the control and ischemia experiment (see above).
3.3. Somatosensory and auditory evoked potentials during tonic pain
Fig. 2. (A) Time-course of the ischemic pain. Mean pain ratings and their standard deviations (n516) during (circles) and immediately after (squares) a 10-min ischemia period (horizontal filled bar). The subjects had to rate the pain every 2.5 min (rating ,4: non painful, rating $4: painful). The deflation of the cuff was followed by immediate relief of pain, which was replaced by post-ischemic paraesthesiae. (B) Phasic pain ratings before, during and after tonic pain (filled symbols). The horizontal dashed line divides non-painful (rating ,4) and painful sensations (rating $4). Note the significant reduction of phasic pain ratings compared to the control session (open symbols). (C) Late somatosensory evoked potentials (peak-to-peak amplitudes). Note the reduction of the SEP-amplitudes during the ischemic pain and the long lasting after-effect. (D) Late auditory evoked potentials (peak-to-peak amplitudes). Note, that there is no significant difference between the AEP-amplitudes from the control and ischemia experiment. Mean values and standard deviations (n516) from control (open symbols) and ischemia experiment (filled symbols). Every symbol stands for one stimulus block. The filled horizontal bar indicates the ischemia duration in the ischemia experiment. Significant differences between the two experimental conditions (two tailed t-test) are marked by asterisks: *** P#0.001; ** P#0.01; * P#0.05.
The effects on the neurophysiological level of measurement were much more distinct (Fig. 2C and Table 1). Like the subjective pain ratings, the SEP-amplitudes showed a habituation across the course of both the control (open symbols) and the ischemia session (filled symbols). But already in the first block during the ischemia there was a significant reduction of these amplitudes by 28% compared with the control experiment. This effect remained in the second half of the ischemia with a reduction by 27% and became still larger after the end of the ischemia period with a reduction by 35%. In the last stimulus block there was a clear increase of the SEP-amplitudes in the ischemia experiment, but the difference between the two experimental conditions remained significant. This long-lasting aftereffect (20 min) of the ischemia shown in the pain related components of the SEP is remarkable. The two-way ANOVA for the AEP showed no effect of tonic pain, only a significant effect of the time course (Table 1). Fig. 2D illustrates that this is again due to a slight habituation, which was similar in the control experiment and under severe tonic pain. Throughout all six stimulus blocks, AEP amplitudes in control and ischemia session were similar. Table 2 Relative changes (post-pre) / pre controlischemia a
ischemia time and the attempt to show a possible aftereffect, the intracutaneous stimuli had to be applied continuously without the break needed for dishabituation. To account for habituation, all statistical tests were performed versus the corresponding time point of the control experiment. Statistical analysis was performed by two-way ANOVA showing a significant effect of the time course
E SEP AEP bEEG a
Control (%)
Ischemia (%)
25.5 213.7 210.0 24.4
226.1 230.6 217.0 148.8
All comparisons were made between block 2 (immediately before ischemia) and block 4 (during maximal ischemic pain).
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
107
Fig. 3. Changes of EEG power during tonic pain. Mean power values and standard deviations (n516) in the classical frequency bands from the control (open bars) and ischemia experiment (filled bars). Every bar stands for one stimulus block. The filled horizontal bar indicates the ischemia duration in the ischemia experiment. (A) Delta (1–3.9 Hz); (B) theta (4–7.4 Hz); (C) alpha 1 (7.5–8.9 Hz); (D) alpha 2 (9–12.4 Hz); (E) beta (12.5–18 Hz). Note the significant increase of beta power during and shortly after ischemic pain. Statistically significant differences between the two experimental conditions (two tailed t-tests) are marked by asterisks: *** P#0.001; ** P#0.01; * P#0.05.
3.4. EEG during tonic pain
4. Discussion
The two-way ANOVA for the power spectrum of the spontaneous EEG showed significant effects only in the beta frequency range. Details are given in Fig. 3: the beta power increased significantly during ischemic pain in comparison with the control experiment. This increase remained during the first ten min after ischemic pain. In the last stimulus block, the beta-power again decreased to control levels. Increases in delta or theta power, which would reflect reduced vigilance, were not observed in this study. To the contrary, there was a small but non-significant decrease around the end of ischemic pain (stimulus blocks 4 and 5). Power in the alpha frequency bands remained remarkably stable throughout the study. Notably, there was also no trend for a shift of power from fast alpha to slow alpha.
The present study examined the counterirritation phenomenon between two painful stimuli using phasic pain, caused by intracutaneous electrical stimuli of one hand, and a concurrent tonic pain caused by ischemia of the other arm. Phasic pain and its neurophysiological correlate, late components of the somatosensory evoked potential, were found to be significantly reduced by heterotopic tonic pain. Pain reduction outlasted the 20-min conditioning stimuli by another 10–20 min. The extent of pain reduction was similar to a regular systemic dose of an opioid drug [32]. On the other hand, neurophysiological control recordings of the auditory evoked potentials and the spontaneous EEG gave no evidence for reduced vigilance or attention during tonic pain. The only EEG effect of the strong tonic pain was a small increase in beta power. The observed after-effect indicates long-lasting changes in the human pain processing systems, triggered by strong tonic pain. After the release of the cuff, there was a sudden relief from ischemic pain, but the subjective ratings of the phasic pain remained reduced for another 10 min. The SEPs showed their largest reduction immediately after the end of the ischemia and remained significantly reduced for 20 min. Similar after-effects of 5–60-min duration were found for the human blink reflex, for the muscular pain threshold [26], for nociceptive reflexes [36] and for heat
3.5. Habituation Table 2 shows the relative changes in all parameters from immediately before ischemia to the time of maximal ischemic pain. The values from the control session document some habituation (around 10%) due to the long lasting sessions with continuous stimulation. But the effects of tonic pain were clearly greater than the effects due to habituation.
108
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
pain ratings [33]. On the other hand, Price and McHaffie [31] reported no after-effect when they reduced the pain sensation of electrical test stimuli by simultaneously applied noxious heat stimuli. Probably, the very short duration (10 s) of their conditioning heat stimuli was responsible for this outcome [15,34,36]. A previous study in our laboratory also failed to document significant aftereffects after 15 min of ischemic pain [14]. But in that study measurements were interrupted for several minutes after each stimulus block to allow for dishabituation, and an after-effect may thus have been missed. In conclusion, with conditioning stimuli of sufficient intensity and duration, long lasting effects on inhibitory pain control systems have regularly been found. The time immediately after release of the cuff is characterized by the appearance of the tingling and pricking post-ischemic paraesthesiae. Microneurographic studies in humans have shown that these paraesthesiae are due to ectopically generated paroxysmal discharges in myelinated fibres; unmyelinated fibres are not involved [25]. This afferent barrage in large myelinated fibres could be involved in the observed after-effects. Pain inhibition by activation of Ab-fibres is the basis of transcutaneous electrical nerve stimulation (classical TENS) and dorsal column stimulation. With these techniques, pain relief is only obtained when the paraesthesiae are projected to the same region as the pain [24,39]. Also the pain relief by vibration is more efficient when applied within the area of pain than when applied to neighbouring areas, contralateral vibration had no effect [5,21]. In our experiments, the post-ischemic paraesthesiae and the phasic test pain were located in contralateral body areas. Thus, local or segmental interactions between large and small diameter fibres can be ruled out as the source of pain inhibition in the postischemia period. The counterirritation phenomenon reported here resembles inhibitory mechanisms in the animal model, termed ‘diffuse noxious inhibitory controls’ (DNIC). Spinal convergent neurons in the rat are inhibited by painful stimulation of body areas outside the receptive field of these neurons [20]. Especially C-fibre stimulation activates these inhibitory control systems, the same fibre population which is excited by ischemic muscle contractions [23]. Long lasting after-effects are characteristic of DNIC. The inhibition is thought to be mediated by a supraspinal loop with a complex pharmacology (for review see Refs. [16,19]). The role of the endogenous opioids in pain control systems is especially well investigated. In animal experiments two different forms of foot shock induced analgesia were found, one opiate mediated and the other without the participation of opiates. Which of both systems will be activated seems to depend on stimulus parameters: The higher the intensity and the longer the duration of the painful stimuli, the more likely it is, that the opiate antagonist naloxone will antagonize the analgesia induced
by exposure to those stimuli [1]. In humans, experiments with naloxone gave contradictory results: Willer et al. [37] demonstrated a naloxone-reversible inhibition of nociceptive withdrawal reflexes during the anticipation of a very strong foot shock. The pain threshold elevation induced by cold pressor pain was naloxone-reversible [18]. The analgesic effect of a long distance run on the perception of ischemic pain was also naloxone-reversible, but in the same experiment, the analgesic effect on noxious thermal stimulation was not [17]. Pertovaara et al. [27] found no effect of naloxone on the increase of the dental pain threshold during ischemic pain. Obviously, also in humans there are different pain control systems which can be activated by noxious stimuli. The effects of tonic pain on the ratings of phasic pain (26% reduction with respect to baseline) met the range of the effects of strong narcotic analgesics. Under similar experimental conditions, tilidine (100 mg, p.o.) decreased phasic pain ratings by 25% in comparison to placebo [9], meperidine (150 mg, p.o.) by 40% [10], and pentazocine (30 mg, i.v.) by 15% [7]. Chapman et al. [12] described decreases in subjective pain report by 23% following fentanyl (0.1 mg, i.v.) and by 25% following nitrous oxide (33% in oxygen). Our control data, however, provide indirect evidence against the involvement of endogenous opioids in the analgesia caused by ischemic pain. Opioids and other centrally acting analgesics affect the spontaneous EEG in a characteristic way: at low doses they slow the dominant alpha frequency, at higher doses they reduce the power in the alpha band and increase the activity in the delta and theta bands [6,11,32]. If endogenous opioids are released during strong tonic pain, we would expect similar changes in the EEG of our subjects. But we found the opposite EEG alterations, a slight decrease of power in the slow frequency range, accompanied by a statistically significant increase of beta power. Similar EEG changes were found during cold pressor pain [3] and in patients suffering from dental pain [13]. In PET studies, ischemic pain did not lead to a displacement of exogenous opioid-receptor ligands (Treede, Campbell, Magiros, and Frost, unpublished data), in contrast to capsaicin-induced pain [4]. These data argue against the action of endogenous opioids at cerebral opioid receptors during ischemic pain. The absence of the typical opioid effects on the EEG during ischemic pain may also indicate that endogenous opioids are only released locally at an appropriate site in the central nervous system, e.g. in the spinal cord dorsal horn. The described EEG alterations gave evidence against a decreased state of vigilance in our subjects during the experiments. Such a decrease of vigilance would lead to a reduction of power in the alpha range and a shift toward lower frequencies [22]. Also the late components of the auditory evoked potentials are strongly influenced by the state of alertness of the subject [29]. We did not observe
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
any significant change in the late components of the AEP during and after strong ischemic pain. In a study on the inhibition of proximal pain perception by distal coldpressor pain, it was argued that the ipsilateral effects were distinct from a change in attention, because they were more pronounced than the contralateral effects [2]. Our data indicate that contralateral pain inhibition is also distinct from distraction. In summary, the inhibition of a phasic test pain by simultaneous tonic pain is not due to altered attention or vigilance, because there were no corresponding changes in AEPs or EEG spectra. The long-lasting after effect (longer than 10 min), which is not explained by segmental mechanisms caused by post-ischemic paraesthesiae, indicates that this inhibition is likely due to the activation of a specific endogenous pain control system. Cerebral release of endogenous opioids, however, would have to be highly localized to this part of the endogenous pain control system, because we found none of the known opioid effects in the spontaneous EEG.
References [1] Z. Amit, Z.H. Galina, Stress-induced analgesia: adaptive pain suppression, Physiol. Rev. 66 (1986) 1091–1120. [2] L. Arendt-Nielsen, K. Gotliebsen, Segmental inhibition of laserevoked brain potentials by ipsi- and contralaterally applied cold pressor pain, Eur. J. Appl. Physiol. 64 (1992) 56–61. [3] M. Backonja, E.W. Howland, J. Wang, J. Smith, M. Salinsky, C.S. Cleeland, Tonic changes in alpha power during immersion of the hand in cold water, Electroenceph. Clin. Neurophysiol. 79 (1991) 192–203. [4] B. Bencherif, M. Oropeza, P.N. Fuchs, R.N. Sheth, S.T. Li, H. Ravert, J. Musachio, B. Mathews, B. Dannals, J.N. Campbell, J.J. Frost, PET scan analysis of m opioid receptor binding during a tonic painful stimulus in normal human subjects, in: Abstracts of 9th World Congress on Pain, Wien, 1999, p. 281. ¨ [5] G. Bini, G. Cruccu, K.-E. Hagbarth, W. Schady, E. Torebjork, Analgesic effect of vibration and cooling on pain induced by intraneural electrical stimulation, Pain 18 (1984) 239–248. [6] T.A. Bowdle, R.J. Ward, Induction of anesthesia with small doses of sufentanil or fentanyl: dose versus EEG response, speed of onset, and thiopental requirement, Anesthesiology 70 (1989) 26–30. [7] B. Bromm, R. Ganzel, W.M. Herrmann, W. Meier, E. Scharein, The analgesic efficacy of flupirtine in comparison to pentazocine and placebo assessed by EEG and subjective pain ratings, Postgrad. Med. J. 63 (1987) 109–112. [8] B. Bromm, W. Meier, The intracutaneous stimulus: a new pain model for algesimetric studies, Methods Find. Exp. Clin. Pharmacol. 6 (1984) 405–410. [9] B. Bromm, W. Meier, E. Scharein, Antagonism between tilidine and naloxone on cerebral potentials and pain ratings in man, Eur. J. Pharmacol. 87 (1983) 431–439. [10] B. Bromm, W. Meier, E. Scharein, Imipramine reduces experimental pain, Pain 25 (1986) 245–257. [11] B. Bromm, W. Meier, E. Scharein, Pre-stimulus / post-stimulus relations in EEG spectra and their modulations by an opioid and an antidepressant, Electroenceph. Clin. Neurophysiol. 73 (1989) 188– 197.
109
[12] C.R. Chapman, Y.M. Colpitts, C. Benedetti, S. Butler, Event-related potential correlates of analgesia; comparison of fentanyl, acupuncture, and nitrous oxide, Pain 14 (1982) 327–337. [13] A.C. Chen, S.F. Dworkin, M.T. Drangsholt, Cortical power spectral analysis of acute pathophysiological pain, Int. J. Neurosci. 18 (1983) 269–278. [14] A.C.N. Chen, R.-D. Treede, B. Bromm, Tonic pain inhibits phasic pain: Evoked cerebral potential correlates in man, Psychiatry Res. 14 (1985) 343–351. [15] J. Ellrich, R.-D. Treede, Characterization of blink reflex interneurons by activation of diffuse noxious inhibitory controls in man, Brain Res. 803 (1998) 161–168. [16] H.L. Fields, A.J. Basbaum, Endogenous pain control mechanisms, in: P.D. Wall, R. Melzack (Eds.), Textbook of Pain, Churchill Livingstone, Edinburgh, 1989, pp. 206–217. [17] M.N. Janal, E.W.D. Colt, W.C. Clark, M. Glusman, Pain sensitivity, mood and plasma endocrine levels in man following long-distance running: effects of naloxone, Pain 19 (1984) 13–25. [18] G. Jungkunz, R.R. Engel, U.G. King, H.J. Kuss, Endogenous opiates increase pain tolerance after stress in humans, Psychiatry Res. 8 (1983) 13–18. [19] D. Le Bars, D. Bouhassira, L. Villanueva, Opioids and diffuse noxious inhibitory control (DNIC) in the rat, in: B. Bromm, J.E. Desmedt (Eds.), Advances in Pain Research and Therapy, Raven Press, New York, 1995, pp. 517–539. [20] D. Le Bars, L. Villanueva, Electrophysiological evidence for the activation of descending inhibitory controls by nociceptive afferent pathways, Prog. Brain Res. 77 (1988) 275–299. ˚ [21] T. Lundeberg, D. Ottoson, S. Hakansson, B.A. Meyerson, Vibratory stimulation for the control of intractable chronic orofacial pain, in: J.J. Bonica, U. Lindblom, A. Iggo (Eds.), Raven Press, New York, 1983, pp. 555–561. [22] M. Matejcek, Vigilance and the EG: psychological, physiological and pharmacological aspects, in: W.M. Herrmann (Ed.), Electroencephalography in Drug Research, Gustav Fischer, Stuttgart, 1982, pp. 405–508. [23] S. Mense, Nociception from skeletal muscle in relation to clinical muscle pain, Pain 54 (1993) 241–289. [24] R.B. North, D.H. Kidd, M. Zahurak, C.S. James, D.M. Long, Spinal cord stimulation for chronic, intractable pain: experience over two decades, Neurosurgery 32 (1993) 384–395. ¨ [25] J.L. Ochoa, H.E. Torebjork, Paraesthesiae from ectopic impulse generation in human sensory nerves, Brain 103 (1980) 835–853. [26] T. Pantaleo, R. Duranti, F. Bellini, Effects of heterotopic ischemic pain on muscular pain threshold and blink relfex in humans, Neurosci. Lett. 85 (1988) 56–60. [27] A. Pertovaara, P. Kemppainen, G. Johansson, S.-L. Karonen, Ischemic pain nonsegmentally produces a predominant reduction of pain and thermal sensitivity in man: a selective role for endogenous opioids, Brain Res. 251 (1982) 83–92. ¨ [28] A. Pertovaara, T. Nurmikko, P.J. Pontinen, Two separate components of pain produced by the submaximal effort tourniquet test, Pain 20 (1984) 53–58. [29] T.W. Picton, S.A. Hillyard, R. Galambos, Habituation and attention in the auditory system, in: W.D. Keidel, W.D. Neff (Eds.), Handbook of Sensory Physiology. Auditory System. Clinical and Special Topics, Springer, Berlin, 1976, pp. 343–389. [30] L. Plaghki, D. Delisle, J.M. Godfraind, Heterotopic nociceptive conditioning stimuli and mental task modulate differently the perception and physiological correlates of short CO 2 laser stimuli, Pain 57 (1994) 181–192. [31] D.D. Price, J.G. McHaffie, Effects of heterotopic conditioning stimuli on first and second pain: a psychophysical evaluation in humans, Pain 34 (1988) 245–252. [32] E. Scharein, B. Bromm, The intracutaneous pain model in the assessment of analgesic efficacy, Pain Rev. 5 (1998) 216–246.
110
A. Reinert et al. / Brain Research 862 (2000) 103 – 110
[33] J.D. Talbot, G.H. Duncan, M.C. Bushnell, M. Boyer, Diffuse noxious inhibitory controls (DNICs): psychophysical evidence in man for intersegmental suppression of noxious heat perception by cold pressor pain, Pain 30 (1987) 221–232. [34] L. Villanueva, D. Le Bars, The encoding of thermal stimuli applied to the tail of the rat by lowering the excitability of trigeminal convergent neurones, Brain Res. 330 (1985) 245–251. [35] S. Watanabe, R. Kakigi, M. Hoshiyama, Y. Kitamura, S. Koyama, M. Shimojo, Effects of noxious cooling of the skin on pain perception in man, J. Neurol. Sci. 135 (1996) 68–73. [36] J.C. Willer, T. De Broucker, D. Le Bars, Encoding of nociceptive
thermal stimuli by diffuse noxious inhibitory controls in humans, J. Neurophysiol. 62 (1989) 1028–1038. [37] J.C. Willer, H. Dehen, J. Cambier, Stress-induced analgesia in humans: endogenous opioids and naloxone-reversible depression of pain reflexes, Science 212 (1981) 689–691. [38] N. Witting, P. Svensson, L. Arendt-Nielsen, T.S. Jensen, Differential effect of painful heterotopic stimulation on capsaicin-induced pain and allodynia, Brain Res. 801 (1998) 206–210. [39] C.J. Woolf, Transcutaneous and implanted nerve stimulation, in: P.D. Wall, R. Melzack (Eds.), Textbook of Pain, Churchill Livingstone, Edinburgh, 1984, pp. 679–690.