Comparative effect of thermal, mechanical, and electrical noxious stimuli on the electroencephalogram of the rat
Institute of Veterinary Animal and Biomedical Sciencesm, College of Sciences, Massey University, Palmerston North, New Zealand
* Corresponding author: Institute of Veterinary Animal and Biomedical Sciences, College of Sciences, Massey University, Postbag 11 222, Palmerston North, New Zealand. E-mail: j.c.murrell{at}massey.ac.nz
Accepted for publication November 17, 2006.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
BACKGROUND: Thermal, mechanical, and electrical stimuli are often used in acute pain studies and cause qualitatively different pain sensations. Yet, the comparative electroencephalogram (EEG) changes caused by these stimuli have not been studied. We hypothesized that because these stimuli cause different pain sensations, EEG responses would also differ.
METHODS: Anaesthesia was maintained with halothane in 46 male Sprague–Dawley rats. The EEG was recorded from the primary somatosensory cortices and vertex. Supramaximal noxious stimuli were applied to the tail and comprised mechanical (forceps clamp 20 N), thermal (52°C water bath), and electrical (50 V, 50 Hz for 2 s) stimuli. The EEG descriptors median frequency (F50), spectral edge frequency (F95), and total power (Ptot) recorded before (baseline) and after noxious stimulation were compared. Data were analysed using two-way factorial ANOVA (stimulus, EEG channel) followed by Bonferroni adjusted post-tests (P < 0.05).
RESULTS: F50 increased during electrical stimulation compared with all baseline periods in all EEG channels, increases from baseline ranging from 115.3 (SD 34.8) to 122.1 (39.6)% for the various channels. A significant increase in F50 during thermal stimulation was identified in some EEG channels, whereas no changes in F50 during mechanical stimulation occurred. Changes in F95 during any stimulus compared with baseline were not significant.
CONCLUSIONS: Different noxious stimuli caused differing EEG changes. As the somatosensory cortex contains relatively few exclusively nociceptive neurons, the EEG recorded from this region during the application of predominantly noxious stimuli (mechanical and thermal) may demonstrate minimal cortical activation compared with non-specific electrical noxious stimuli.
Keywords: anaesthetics volatile, halothane; brain, electroencephalography; pain, experimental; pain, mechanism;
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Noxious stimuli cause changes in the electroencephalogram (EEG).1 The typical EEG response to noxious stimuli is desynchronization;2 a shift towards high-frequency, low-amplitude activity. This change in activity is thought to represent the cerebral processing of noxious stimuli, which would be associated with pain in conscious individuals. Less commonly, synchronization is reported: a shift in the EEG towards high-amplitude, low-frequency activity;3 however, other clinical studies have failed to identify EEG changes in response to noxious stimulation. Therefore, current evidence indicates that the EEG can be a robust tool to measure nociception, but only in rigorously controlled experimental situations.
Thermal, mechanical, and electrical stimuli are commonly used in cutaneous phasic pain studies in people and cause qualitatively different pain sensations.4 The mechanisms that allow pain from cutaneous supramaximal thermal, mechanical, and electrical noxious stimuli to be distinguished are unknown. The cross specificity of most nociceptors suggests that the encoding of information leading to pain perception is unlikely to be determined solely by peripheral nociceptor afferent input to the central nervous system (CNS).
A number of EEG studies have investigated the effect of different anaesthetic agents on phasic noxious mechanical, thermal, or electrical stimuli, and these stimuli are commonly used in models investigating the modulation of EEG changes by anaesthetic drugs. However, the comparative effects on EEG by thermal, mechanical, and electrical noxious stimuli have not been previously investigated. This knowledge is important to allow valid comparison of data among studies using different stimulus modalities.
There were two major aims of this study. Firstly, to compare the EEG changes caused by cutaneous application of supramaximal noxious thermal, mechanical, and electrical stimuli. We hypothesized that these stimuli would cause quantitatively different EEG responses, because the perception of pain associated with each stimulus is qualitatively different. Secondly, to identify the most promising modality of noxious stimulation that could be applied to future studies investigating EEG changes during the application of a cutaneous phasic acute pain stimulus.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
This study was approved by the Massey University Animal Ethics Committee. Forty-eight male Sprague–Dawley rats weighing 379 (25) g were studied. Anesthesia was induced and maintained with halothane. An 18-gauge cannula was placed in the trachea via the mouth and ventilation was controlled using IPPV, adjusted to maintain end-tidal carbon dioxide concentration (E'Co2) between 4.7 and 6.0 kPa. Monitoring comprised end-tidal halothane concentration (E'Hal) and E'Co2 using a side-stream gas analyser (Hewlett Packard M1025B gas monitor). Rectal temperature was monitored continuously using a digital thermometer and maintained between 37.0 and 38.5°C with a circulating warm water blanket heating device. The whole tail of the rat was placed in the centre of a heating coil, built in-house from an infusion extension set. Water at 29ºC was continuously circulated through the coil using a pump and temperature-controlled water bath. The coil was briefly removed from the tail to allow the application of noxious stimuli.
The head of the rat was secured in a stereotaxic apparatus and the skull periosteum exposed by a midline skin incision. Six silver–silver chloride electrodes (0.4 mm diameter x 8 mm Ag–AgCl segment, In Vivo Metric) were placed in contact with the dura through holes drilled in the skull. Two active electrodes were placed over the left and right primary somatosensory cortices at a loci corresponding to afferent sensory input from the tail (S1R, S1L) (2.5 mm caudal to bregma, 2.5 mm left and right of midline).5 Two other active electrodes were placed over the left and right vertices (VxL, VxR) (4.5 mm caudal to bregma, 1.5 mm left and right of midline).6 The reference electrode was placed over the fontal sinus (10 mm cranial to bregma, 1 mm left of midline) and a ground electrode was placed in the right caudal quadrant (7 mm caudal to bregma, 2.5 mm right of midline).
The EEG was recorded continuously from the left and right somatosensory cortices and vertices. Signals from each channel of EEG were fed via custom built ‘break-out’ boxes to separate but identical amplifiers (Iso-Dam isolated biological amplifier, World Precision Instruments Inc). The signals were amplified with a gain of 100 Hz and a band pass of 0.1–100 Hz, and recorded on a Powerlab 4/20 data acquisition system (AD Instruments Ltd), which digitized the signal online at a rate of 1.0 kHz. The digitized signal was recorded on an Apple Macintosh personal computer.
All noxious stimuli were applied to the tail of the rat. The electrical stimulus was given to the right lateral tail base via two stainless steel needle electrodes inserted through the epidermis and spaced 1 cm apart. The electrical stimulus was generated by a stimulator (S48K Square Pulse Stimulator, Astro-Med Inc., Grass Instrument Division) and comprised a 50 V, 50 Hz stimulus applied for 2 s. The impedance of the electrodes in situ was 4 k
.
The mechanical stimulus was also applied to the base of the tail. A pair of artery forceps was modified to deliver a force of 20 N when the base of the tail was placed between the arms of the forceps and the forceps closed until the 1st ratchet locked. The force was applied for 5 s. The thermal stimulus comprised a 55°C water bath in which the whole tail was immersed for 5 s.
Rats were randomly divided into three groups to receive either the electrical, mechanical, or thermal stimulus first. The order of presentation of the second and third stimuli was also randomly allocated within each of the three groups. After completion of electrode placement, the E'Hal was stabilized at 1 (0.05)% for 10 min before application of any stimuli and kept constant for the rest of the experiment. Each stimulus was applied once, with a 10-min rest period between stimuli. Two minutes of baseline EEG data were recorded before stimulation, followed by 2 min of EEG data immediately after stimulation.
Raw data from the EEG were inspected manually and any artifacts were excluded from further analysis. The total power (Ptot), median frequency (F50), and spectral edge frequency (F95) for 1-s epochs were calculated using specialized software (FFT program Craig Johnson). Data from each channel of the EEG recording were analysed separately. For the purposes of statistical analysis, EEG data from the first 30 s of each baseline period and the first 30 s of data recorded after noxious stimulation were compared. A single mean value for each EEG parameter was calculated for each of these time periods in every rat.
Statistical analysis
Statistical analysis of data was carried out using Prism 4 for Macintosh Version 4b (Graphpad Software Inc.). Statistical significance was assumed when P < 0.05. Data were normally distributed (Kolmogor–Smirnov test). Preliminary analyses proved that there were no statistically significant differences between the three baseline periods and that the order of stimulus presentation did not influence the percentage change in F50. Therefore, data pertaining to all three stimuli from the three groups of rats were pooled. A Model 1 two-way analysis of variance for F50, F95, and Ptot was carried out, with stimulus (electrical, mechanical, thermal, and baseline) and channel (S1R, S1L, VxR, and VxL) as fixed variables. Bonferroni post-tests for multiple comparisons were carried out when P < 0.05, comparing EEG recorded during the baseline and noxious stimulation periods to identify where differences lay.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Complete data sets were collected from 42 animals. Of these animals, 39 demonstrated purposeful complex movement responses during the application of all noxious stimuli comprising hind-limb withdrawal and tail movements. These responses were of short duration and stopped when stimulation ceased.
For F50 interaction between the type of stimulus and EEG channel was not significant (F15,800 = 0.26, P = 0.9978). Stimulus (baseline 1, baseline 2, baseline 3 or thermal, electrical or mechanical) had a significant effect on the variance in F50 (F5,800 = 22.91, P < 0.0001). EEG channel also had a significant effect, although smaller effect than stimulus on the variance in F50 (F3,800 = 2.75, P = 0.047). F50 increased significantly (P < 0.05) during electrical stimulation compared with all baseline periods in all channels (Table 1). F50 increased during thermal stimulation compared with the first and third baseline periods in the S1R and VxR channels only (P < 0.05) (Table 1). There were no changes in F50 during mechanical stimulation compared with the baseline.
|
Similar to F50, for Ptot, the interaction between type of stimulus and EEG channel was not significant (F15,800 = 0.61, P = 0.869). Stimulus had a significant effect on variance in Ptot (F5,800 = 6.32, P < 0.001). EEG channel also significantly influenced Ptot variance, although similar to F50, this influence was smaller in magnitude than the stimulus effect (F3,800 = 4.80, P = 0.0031). Ptot decreased during the application of an electrical stimulus compared with two of the baseline periods (baselines 2 and 3) in the S1R EEG channel only (Table 2). A decrease in Ptot also occurred during mechanical stimulation compared with baselines 2 and 3 in this channel (Table 2). There were no changes in Ptot during thermal stimulation compared with baselines.
|
There were no significant changes in F95 during application of any noxious stimulus compared with any of the baseline periods.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
This is the first study to compare the effects on EEG by supramaximal cutaneous thermal, mechanical, and electrical stimuli and demonstrates that the EEG changes caused by these stimuli are quantitatively different from each other. The electrical stimulus caused a robust increase in F50 compared with all baseline periods in all EEG recording channels. Although F50 increased after thermal stimulation compared with two baseline periods, the magnitude of this change was less after thermal compared with electrical stimulation. Increases in F50 are characteristic of a desynchronization or arousal reaction to noxious stimulation and have been widely reported in other studies,7–10 reflecting a switch towards high-frequency, low-amplitude EEG activity.
This experiment used the minimal anaesthesia model paradigm11 to investigate the effects on EEG by noxious stimulation. The cerebral cortex remains responsive to noxious stimuli at a light plane of halothane anaesthesia,12 thus providing an ethical model in which to examine the EEG effects of stimuli which would be painful in a conscious individual. On the basis of published analgesiometry studies, the three stimuli were considered to be supramaximal, sufficient to cause pain in an awake animal. The minimal anaesthesia model has been applied to other species to investigate EEG changes during the application of simple clinical noxious stimuli such as castration in lambs and horses and velvet antler removal in deer.8–10 These interventions were associated with an increase in F50 and therefore it was anticipated that the thermal, mechanical, and electrical stimuli in the present investigation would also cause an increase in F50.
A number of functional magnetic resonance imaging (fMRI) techniques have investigated changes in brain activation during noxious stimulation in both humans and animals. Disbrow and collegues13 compared the response of the somatosensory cortex to noxious mechanical, thermal, and electrical stimuli in humans using fMRI. No cortical activation was found in response to noxious heat or mechanical stimulation, a finding supported by others during several positron emission topographic studies of cerebral flow during noxious stimulation.14 15 The somatosensory cortex is composed of relatively few nociceptive specific neurons, and these are interspersed with non-nociceptive neurons.16 Therefore, it is suggested that pure nociceptive stimuli such as thermal or mechanical cause little or no S1 activation in imaging studies because of the intermixed organization of the responding fibres in the cortex. A similar principle can be used to explain the relative quiescence of the EEG recorded over S1 during mechanical and thermal stimuli compared with the electrical stimulus. An electrical stimulus indiscriminately activates all local receptors, both nociceptive and non-nociceptive, and therefore resulted in greater activation of S1 neurons, causing an increase in F50.
The results of the present study are disparate to the findings of others17 18 where mechanical and electrical stimuli caused similar increase in F50. However, in these studies mechanical stimulation of the tail comprised application and oscillation of a tail clamp at 1 Hz. It is possible that the oscillating clamp, compared with a stationary forceps clamp, was associated with greater activation of non-nociceptive neurons, resulting in an increase in F50.
When noxious thermal stimulation did cause an increase in F50 in the present investigation, this activation was lateralized to right-sided S1 and Vx electrodes. The thermal stimulus was applied to the whole length of the tail therefore this lateralization is difficult to rationalize. Studies of cerebral lateralization in man have shown that the right cerebral hemisphere is more involved in spatial and in attentional processes and many aspects of emotion.19 20 This has led to the theory of right hemisphere ‘conservatism’,21 whereby threatening external stimuli that are of great importance for survival are directed towards the right hemisphere. Pain integration, a determinant factor in species survival, may therefore also preferentially activate the right cerebral hemisphere. Others have also identified a predominance of activation in the right cerebral hemisphere during painful stimulation and this may account for the activation in the right-sided EEG channels in the present investigation.
A reduction in Ptot was found during electrical and mechanical stimulation in the S1R EEG channel, however in contrast to F50, no change during thermal stimulation was identified. A decrease in Ptot has been found in other studies investigating EEG changes and nociception in the minimal anaesthesia model,9 10 22 although this is not a universal finding.8 However, there is a trend for an EEG desynchronization response to also be associated with a reduction in total power.22 23 It has been proposed that changes in Ptot and F50 may reflect different aspects of noxious stimulation.10 24 The reduction in Ptot may be associated with a decrease in adequacy of anaesthesia caused by the stimulus, whereas the increase in F50 reflects the noxious element of the stimulus. This may explain the discrepancy between changes in F50 and Ptot for the different stimulus modalities.
The absence of changes in F95 during noxious stimulation using any modality supports the findings of previous studies using the minimal anaesthesia model,8–10 suggesting that F95 is an insensitive indicator of nociception.
Recording electrodes were placed bilaterally over the primary somatosensory cortex (S1) in the somatotopic region representing the tail5 and the vertex.25 In awake rats, somatosensory evoked potentials (SEPs) recorded from the vertex location compared with the S1 show increased sensitivity to hypnotic and analgesic drugs.6 25 It has been proposed that although SEPs recorded from both locations are related to functional pain mechanisms, the vertex SEP signals pain unpleasantness, while the S1 SEP relates to spatio-temporal localization and intensity quantification of noxious stimuli.26 Although the animals in the present study were anaesthetized and therefore unable to experience pain, we recorded from both loci in order to investigate whether the EEG responses at each loci were different to each other. EEG changes were more commonly identified from S1 recording electrodes than the vertex. The animals in the present study were anaesthetized, therefore neural activity related to the ‘emotional colour’ of pain or pain unpleasantness is likely to be greatly diminished compared with the awake state.
All noxious stimuli caused complex movement responses in the majority of animals, indicating that they were of similar potency. Procedures were adopted to ensure that noxious stimuli were standardized. The tail was maintained constant in a warm water coil to prevent a decrease in tail cutaneous temperature because of anaesthesia, as a fall in tail skin temperature may alter the EEG response to thermal stimulation of the tail. The tail was always stimulated in the same place for the respective stimuli and the E'Hal concentration was tightly controlled between animals. The response of the CNS to noxious input is recognized to change as a result of prior activation,27 therefore the order of application of the different stimuli was randomized. We found no evidence of hypersensitivity to noxious stimulation developing during the course of the three different stimuli and there was no change in the baseline EEG data recorded between stimuli. Aberrations in physiological parameters such as body temperature and carbon dioxide concentration can also influence the EEG, but these were controlled within normal physiological limits in all animals.
Electrical noxious stimulation produced the most reliable and repeatable EEG response, with a significant increase in F50 during electrical stimulation compared with all baseline periods. Electrical stimuli are easy to control, quantifiable, and reproducible and are therefore commonly used in nociception studies in humans and animals. However, there are also disadvantages associated with this modality. Transcutaneous electrical stimulation results in non-selective direct activation of all peripheral afferent fibres, at the same time bypassing the transduction part of the nociceptive pathway. The afferent pathway is also activated in a synchronized and unnatural manner. Therefore, although the electrical stimulus used in the present study would have caused pain in a conscious animal, undoubtedly it would have resulted in both noxious and non-noxious input to the cerebral cortex.
Thermal, mechanical, or electrical phasic noxious stimuli are frequently applied in studies investigating EEG responses to noxious stimulation in animals, and these models have been used to investigate modulation of these responses by co-administration of different drugs. The results of the present study emphasize that these different noxious stimulus modalities cause differing EEG responses and therefore extrapolation of data between EEG studies using different modalities should be avoided. Electrical stimulation caused the most reliable and repeatable increase in F50, an EEG change that is indicative of noxious stimulation. The variable EEG response to the mechanical and thermal stimuli used in the present study suggest that these are unsuitable stimuli to investigate EEG responses to noxious stimulation in anaesthetized rats.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
1 Bromm B. (1985) Modern techniques to measure pain in healthy man. Methods Find Exp Clin Pharmacol 7:161–9.[Web of Science][Medline]
2 Arduini A and Arduini M. (1954) Effects of drugs and metabolic alterations on brain arousal. J Pharmacol Exp Ther 110:76–85.
3 Prince D and Shanzer S. (1966) Effects of anaesthetics upon the EEG response to reticular stimulation. Patterns of slow synchrony. Electroen Clin Neuro 21:578–88.[CrossRef]
4 Arndt J and Klement W. (1991) Pain evoked by polymodal stimulation of hand veins in humans. J Physiol 440:467–78.
5 Chapin J and Lin C. (1984) Mapping the body representation in the SI cortex of awake and anaesthetised rats. J Comp Neurol 229:199–213.[CrossRef][Web of Science][Medline]
6 Stienen P, De Groot H, Venker van Haagen A, van den Brom W, Hellebrekers L. (2005) Differences between somatosensory-evoked potentials recorded from the posteriolateral thalamic nucleus, primary somatosensory cortex and vertex in the rat. Brain Res Bull 67:269–80.[CrossRef][Web of Science][Medline]
7 Ichinose F, Miyazaki M, Goto T, et al. (1999) Electroencephalographic responses to the formalin test in rats. Pain 80:251–6.[CrossRef][Web of Science][Medline]
8 Johnson C, Stafford K, Slyvester S, Ward N, Mitchinson S, Mellor D. (2005) Effects of age on the electroencephalographic response to castration in lambs anaesthetised using halothane in oxygen. New Zeal Vet J 53:433–7.
9 Johnson C, Wilson P, Woodbury M, Caulkett N. (2005) Comparison of analgesic techniques for antler removal in halothane-anaesthetized red deer (Cervus elaphus). Vet Anaesth Analg 32:61–71.[CrossRef][Web of Science][Medline]
10 Murrell J, Johnson C, White K, Taylor P, Haberham Z, Waterman-Pearson AE. (2003) Changes in the EEG during castration in horses and ponies anaesthetized with halothane. Vet Anaesth Analg 30:138–46.[CrossRef][Web of Science][Medline]
11 Murrell J and Johnson C. (2006) Neurophysiological techniques to assess pain in animals. J Vet Pharmacol Ther 29:325–35.[Medline]
12 Barter L, Dominguez C, Cartsens E, Antognini J. (2005) The effect of isoflurane and halothane on electroencephalographic activation elicited by repetitive noxious c-fiber stimulation. Neurosci Lett 382:242–7.[CrossRef][Web of Science][Medline]
13 Disbrow E, Bounocore M, Antognini J. (1998) Somatosensory cortex: A comparison of the response to noxious thermal, mechanical and electrical stimuli using functional magenetic resonance imaging. Hum Brain Mapp 6:150–9.[CrossRef][Web of Science][Medline]
14 Casey K, Minoshima S, Berger K, Koeppe R, Morrow T, Frey K. (1994) Positron emission topographic analysis of cerebral structures activated specifically by repetitive noxious heat stimuli. J Neurophysiol 71:802–7.
15 Talbot J, Marrett S, Evans A, Meyer E, Bushnell M, Duncan G. (1991) Multiple representations of pain in the human cerebral cortex. Science 251:1355–8.
16 Casey K and Morrow T. (1983) Nocifensive responses to cutaneous thermal stimuli in the cat: Stimulus-response profiles, latencies and afferent activity. J Neurophysiol 50:1497–515.
17 Dominguez C, Barter L, Antognini J. (2005) Intrathecal picrotoxin minimally alters electroencephalographic responses to noxious stimulation during halothane and isoflurane anaesthesia. Acta Anaesthesiol Scand 49:763–70.[CrossRef][Web of Science][Medline]
18 Orth M, Barter L, Dominguez C, Atherley R, Carstens E, Antognini J. (2005) Halothane and propofol differentially affect electroencephalographic responses to noxious stimulation. Br J Anaesth 95:477–84.
19 Gainotti G. (1996) Lateralization of brain mechanisms underlying autonomic and controlled forms of spatial orienting and attention. Neurosci Biobehav Rev 20:617–22.[CrossRef][Web of Science][Medline]
20 Silberman E and Weingartner H. (1986) Hemispheric lateralization of functions related to emotion. Brain Cogn 5:322–53.[CrossRef][Web of Science][Medline]
21 Geschwind N and Galaburda A. (1985) Cerebral lateralization. Biological mechanisms, associations and pathology: I. A hypothesis and a program for research. Arch Neurol 42:428–59.
22 Antognini J, Carstens E, Sudo M, Sudo S. (2000) Isoflurane depresses electroencephalographic and medial thalamic responses to noxious stimulation via and indirect spinal action. Anaesth Analg 91:1282–8.
23 Chen A, Dworkin S, Drangsholt M. (1983) Cortical power spectral analysis of acute pathophysiological pain. Int J Neurosci 18:269–78.[Web of Science][Medline]
24 Traast HS and Kalkman CJ. (1995) Electroencephalographic characteristics of emergence from propofol/sufentanil total intravenous anaesthesia. Anaesth and Analg 81:366–71.
25 Stienen P, van den Brom W, de Groot H, Venker van Haagen A, Hellebrekers L. (2004) Differences between primary somatosensory cortex- and vertex-derived somatosensory-evoked potentials in the rat. Brain Res 1030:256–66.[CrossRef][Web of Science][Medline]
26 Stienen P, van Oostrom H, van den Bos R, de Groot H, Hellebrekers L. (2006) Vertex-recorded, rather than primary somatosensory cortex-recorded, somatosensory-evoked potentials signal unpleasantness of noxious stimuli in the rat. Brain Res Bull 70:203–12.[CrossRef][Web of Science][Medline]
27 Woolf C. (1996) Windup and central sensitization are not equivalent. Pain 66:105–8.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||