A study on the effects of thermal pain stimulation at the oral mucosa in EEG

doi:10.21203/rs.3.rs-2234020/v1

Download PDF

Research Article

A study on the effects of thermal pain stimulation at the oral mucosa in EEG

https://doi.org/10.21203/rs.3.rs-2234020/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objectives

The aim of this study was to investigate how thermal pain stimulation on the oral mucosa affects the brain’s pain perception and brain network connectivity using electroencephalograms (EEGs) and to reveal the central systemic mechanisms of oral chronic pain such as burning mouth syndrome (BMS).

Materials and methods

EEG data were collected from 40 healthy volunteers (24 males and 16 females) at rest and during thermal stimulation of the oral mucosa at 45°C, 48°C, 51°C, and 54°C. The power of six frequency bands was calculated for the acquired EEG data. A directed transform function (DTF) was used to construct the network and calculate the network connection parameters.

Results

The results showed that subjective pain scores increased significantly with increasing temperature (p < 0.001). Compared with the resting baseline state, the average power of both theta and alpha bands decreased with increasing temperature (p < 0.001), while the average power of the gamma band increased (p < 0.001). Moreover, the mean values decreased, as did the clustering coefficients and global efficiency in the low and middle frequency bands.

Conclusions

The thermal stimulation of the oral mucosa mainly affects the alpha power decreasing in the contralateral somatosensory area.

Clinical Relevance

This experiment provides a theoretical basis and technical support for studying the mechanism of transmission and perception of oral thermal pain through the trigeminal pathway to the central nervous system in human subjects and provides a theoretical basis for the central system mechanism of chronic pain syndromes such as BMS.

electroencephalogram

Granger causality

oral mucosa

thermal stimulation

Temperature sensation is one of the basic senses of the human body, which not only allows the human body to feel changes in the temperature of the surrounding environment, but also to engage in self-protection. Thermal stimuli that exceed a person's thermal threshold can cause pain. The oral cavity is the initial part of the digestive system, and it plays an important role in protection, perception, and secretion. Cold and warm receptors in the oral cavity are mainly located in the gums and lips. Research has shown that there are fewer warm spots than cold spots on the oral mucosa and they are irregularly distributed, so it is possible that the oral mucosa is more susceptible to damage from high temperatures [2]. Physiological studies have confirmed that the oral mucosa is innervated by the trigeminal nerve [3, 4]. The trigeminal nerve provides afferent information from the skin, sinuses, mucous membranes, and the oral cavity about temperature, touch, and pain [5] and transmits this information to the cerebral cortex.

Burning mouth syndrome (BMS) was classified under the heading “painful cranial neuropathies” by the International Headache Society [6]. Previous studies have suggested that there are changes in the pain modulation system in the brain of patients with BMS [7]. Albuquerque et al. reported that patients with BMS showed significant activation in brain regions such as the precuneus and precentral gyrus during painful stimulation of the trigeminal nerve. Those studies suggested that neuropathic pain in the central nervous system (CNS) is one of the mechanisms of BMS.

A number of studies have investigated electroencephalogram (EEG) responses to various thermal painful stimuli. For example, Dounia et al. found that exposure to both hot and cold stimuli elicited periodic EEG responses and perception [8]. The response to cold stimuli was smaller in magnitude and shorter in latency compared to the response to warm stimuli. Peng et al. studied changes in cortical activity during tonic thermal perception and investigated the effect of directed/distracted attention on these responses. They observed that contralateral-central alpha inhibition may primarily reflect top-down cognitive processes, such as attention [9]. Zhang et al. explored changes in sympathetic skin response (SSR) and EEG under thermal stimuli and sketched a topographic map of the SSR threshold temperature of the oral mucosa. Their results revealed that the autonomic and central nervous system played important roles in the thermal perception of the oral mucosa and could be helpful for better understanding the pathological mechanism of BMS. Therefore, it is of interest to study the changes and mechanisms of CNS responses in the oral mucosa under different thermal stimulation conditions.

In this study, we investigated the effect of thermal stimulation on the rhythmic activity of the brain in healthy subjects by applying thermal stimuli of different temperatures to the oral mucosa. We investigated the main frequency ranges, power, and brain regions of the whole brain most closely associated with thermal stimulation of the oral mucosa using EEG data obtained from healthy subjects before and after thermal stimulation. Furthermore, we constructed Granger causal networks based on a directed transform function (DTF) matrix and explored the effects of thermal painful stimulation on the human cerebral cortex from the network connectivity viewpoint. We expected to find brain changes associated with the central mechanisms of BMS.

2.1 Participants

Forty healthy subjects from Tianjin Medical University participated in this study, including 24 males and 16 females, age 19 to 28 years old (mean age ± SD: 23 ± 3, five were left-handed) (Supplementary Table 1). All the participants had complete dentition without any mucosal diseases or abnormal temperature sensations in the oral cavity. None of the participants had any neurological disease or consumed psychotropics or drugs that would inhibit glandular secretion in the past two weeks, such as diazepam and atropine. This study was approved by the ethics committee of Tianjin Medical University (No. TMUhMEC20210201) and was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki. Informed consent was obtained before the study, and the subjects were able to withdraw from this study at any time.

2.2 the Instrumentation of the Experiment

The thermal stimulus instrument worked based on the principle of intelligent temperature control (Chinese Patent No. ZL202121087859.2) (Supplementary Fig. 1). Thermal stimuli were applied on an area of 5 x 5 mm through a Peltier semiconductor chip. Real-time temperature stability was determined by a PT100 sensor (Heraeus, Germany) temperature control system (Supplementary Fig. 2). Real-time data from the sensor was sent to the temperature control computer where fluctuations of ± 0.5°C were recorded, while. EEG data were acquired by a 32-channel Quik-Cap through a NuAmps amplifier (Compumedics Neuroscan, Australia). Data were referenced to the left mastoid with a ground at electrode Fpz. The EEG electrodes were placed according to the extended international 10–20 system. Impedance levels were set at less than 5 kΩ. The sampling frequency was set to 1000 Hz.

2.3 Experimental Paradigm

The thermal stimulus test was conducted in the morning every day. Before the experiment, each participant was told to stay relaxed and not to have a particular thought other than to pay attention to the stimulus. Because the incisor area is the most sensitive area to temperature, the thermal stimuli sites were selected as the central areas of the maxillary and mandibular incisor areas [2]. In a pilot study, 36°C was used as the starting point for thermal stimulation and the temperature was gradually increased while subjects reported their subjective sensations. When the stimulus temperature reached 45°C, subjects reported being able to slightly feel the temperature of the stimulus. The pilot study gradually increased the temperature at intervals of 1°C., 2°C., 3°C., and 4°C. It found the subjects did not subjectively perceive a significant difference in temperature stimulation at 1°C and 2°C intervals. The reason for that may be that the temperature change was small and did not exceed the oral temperature adaptive temperature threshold. In contrast, when the interval was more than 3°C, the subjects were able to perceive a more substantial temperature change. Therefore, in this study, 45°C was used as the starting stimulation temperature and 3°C was used as the temperature stimulation increment. The four thermal stimuli were 45°C, 48°C, 51°C, and 54°C, each of which had a duration of 8 s. The thermal chip was moved slightly after each recording to avoid sensitization or habituation at the same stimulus site. After the stimulation of the maxillary incisor and mandibular incisor at the same temperature, the stimulus at the next temperature was presented. After each stimulus, the subjects were asked to rate its level of pain on the Visual Analogue Scale/Score (VAS), with 0 indicating no pain, and 10 indicating severe pain, with other ratings indicating different degrees of pain.

2.4 Data Analysis

Preprocessing. The original acquired EEG signals were mixed with background noise, power frequency interference, and other artifacts. Therefore, the EEGLAB toolbox (version 13.6.5b) (Delorme and Makeig, 2004) of MATLAB R2016b (The MathWorks, Natick, MA, USA) was used to conduct baseline correction and remove baseline drift of EEG data. A band-pass filter of 0.5 Hz to 80 Hz was used to filter the data. The 50 Hz power frequency interference was filtered out with a trap of 49.5–50.5 Hz. Independent component analysis (ICA) was used to remove eye movement and muscle artifact components. The pre-processed data were used for subsequent analyses.

Frequency Analysis. Frequency analysis was performed using the periodogram method in MATLAB [10]. Then, the average power of six frequency bands were calculated: delta (1–4 Hz), theta (4–8 Hz), alpha (8–13 Hz), low beta (13–20 Hz), high beta (20–30 Hz), and gamma (30–45 Hz). The trends of the average power of the EEG delta, theta, alpha, low beta, high beta, and gamma bands were analyzed in accordance with the temperatures of the different thermal stimuli applied to the oral mucosa. The electrodes were selected and grouped into five main regions to compute global frequency changes, based on previous studies: the frontal, central, left parietal, right parietal, and occipital areas [11]. The frontal region included the F3, Fz, F4, Fc3, Fcz, and Fc4 electrodes, the central region included the Cz, Cpz, and Pz electrodes, the left parietal region included the C3, Cp3, and P3 electrodes, the right parietal region included the C4, Cp4, and P4 electrodes, and the occipital region included the O1, OZ, and O2 electrodes (Fig. 1).

Directed Transform Function. Directed Transform Function (DTF) is a physical quantity that describes the direction and strength of information transfer between the signals of each channel. The strength of the causal relationship between channels can be judged by the size of the DTF value, as proposed by Kaminski and Blinowska [12]. It is a method to quantitatively characterize causal connections of frequency domains based on a multivariate autoregression (MVAR) model, and it extends Granger causality analysis to variables of any dimension [13]. Studies have confirmed that using DTF methods to assess the connectivity patterns of the cerebral cortex is effective [14]. The causal connection strength ${\gamma }_{ij}^{2}\left(f\right)$ from channel $j$ to channel$i$can be expressed as:

$$\gamma _{{ij}}^{2}\left( f \right)={\left| {{H_{ij}}\left( f \right)} \right|^2}/\sum\limits_{{m=1}}^{k} {{{\left| {{H_{im}}\left( f \right)} \right|}^2}}$$

In the formula, $H\left(f\right)$ is the transfer matrix and ${\gamma }_{ij}^{2}\left(f\right)$ is the normalized DTF matrix. In this paper, the calculation of DTF was determined by eConnectome toolbox developed by He et al. [15].

${DTF}_{mean}$ . This is the average value of the DTF matrix that is used as an indicator that directly describes the causal connection strength of the EEGs. In the causal network, ${DTF}_{mean}$ is defined as the ratio of the actual number of connected edges to the possible total number of connected edges [16]. The calculation formula is as follows:

$$DT{F_{mean}}=\frac{1}{{k\left( {k - 1} \right)}}\sum\limits_{{i \in K}} {} \sum\limits_{{j \ne i \in K}} {DT{F_{ij}}}$$

In the formula, ${DTF}_{ij}$ represents the directional transfer function matrix from channel $j$ to channel $i$, $k$ is the number of nodes, $K$ is the set of all nodes in the causal network, and the value of ${DTF}_{ij}$ is the edge of the causal network. Higher values of the ${DTF}_{mean}$ indicate there are more node connections, and it also reveals the degree to which the nodes coordinate activities, from a global point of view, in the causal network.

Graph Theory. Graph Theory is a mathematical theory and method for quantitative research and analysis of network topology. Complex network analysis based on graph theory can mathematically quantify the changes in topological properties of network structures; hence, it is widely used in the research on brain networks. It is necessary to set a threshold to transform the DTF matrix into a binary directed matrix to establish a functional network using causal analysis as a measure of network connectivity. When a threshold (T) is given, all the elements greater than T in the DTF matrix are set to 1, indicating that there is a causal connection between the corresponding nodes; all elements less than T are set to 0, indicating that there is no connection between the corresponding nodes. Without causality of the node itself, all diagonal elements of the connection matrix are set to 0.

The clustering coefficient and global efficiency are two important parameters of small-world network properties and are mainly considered in this study. The clustering coefficient (C) is a physical quantity that represents the degree of closeness between nodes and adjacent nodes in the network, and it is an index to measure the degree of grouping of the network [17]. The calculation formula is:

$C=\frac{1}{N}\sum\limits_{{i \in K}} {{c_i}}$ , ${c_i}=\frac{{2{e_i}}}{{{k_i}({k_i} - 1)}}$ (3)

where ${c}_{i}$ is the clustering coefficient of node$i$, ${e}_{i}$ is the number of adjacent actual connected edges of node $i$, and ${k}_{i}$ is the degree of node $i$. In the network, the average of the clustering coefficients of all networks is the clustering coefficient $C$ of the network.

Global efficiency (E) reflects the global transmission capability of the network [18] and measures the comprehensive index of the speed of information transmission in the network. In a network, the higher the global efficiency, the faster the rate of information transfer between nodes. The formula for calculating global efficiency is:

$$E{\text{=}}\frac{1}{{N(N - 1)}}\sum\limits_{{i,j \in V,i \ne j}} {\frac{1}{{{l_{ij}}}}}$$

where ${l}_{ij}$ is the characteristic path length from node $i$ to node $j$.

2.5 Statistical Analysis

Statistical analysis was performed using SPSS statistics version 25.0 (SPSS, Chicago, IL, United States). All the data were tested for normality using the Shapiro–Wilk normality test. The statistical indicators used in this paper were average power, the $DTF$_mean, the clustering coefficient, and the global efficiency of the baseline and the four temperature stimuli. One-way repeated-measures analysis of variance (ANOVA) was used to analyze the statistical differences in the corresponding indicators between baseline, 45° C, 48° C, 51° C, and 54° C stimuli. The Bonferroni correction was used to perform post-hoc two-by-two multiple comparisons. The P-values were for the levels of statistical significance were: * p < 0.05, ** p < 0.01, and *** p < 0.001.

3.1 Results of the Frequency Analysis

The average power of each of the six frequency bands in five brain regions was calculated under the five conditions: baseline, 45°C, 48°C, 51°C, and 54°C thermal stimuli (Fig. 2). Table 1 shows the results of the one-way repeated-measures ANOVA for average power. In the delta band, the average power under the five conditions was statistically different in the frontal (F_{4, 76} = 2.588, p = 0.043), central (F_{4, 76} = 2.657, p = 0.039), and right parietal regions (F_{2.8, 53.201} = 4.904, p = 0.005). The post-hoc two-by-two multiple comparisons (Supplementary A. Table 2) showed there was no significant difference between the average power under the five conditions, except for 48°C and 54°C (p = 0.008) in the right parietal region. In the theta band, a global reduction was observed in all five regions with increasing temperature. In all five regions, the average power showed a significant reduction under the four thermal stimulus conditions compared to baseline. Significant differences in average power were also found between the 45°C and 54°C conditions (p < 0.001) and at 48°C and 54°C (p = 0.029) in the right parietal region.The average power of the alpha band showed the same trend and a similar level of statistical significance as the theta band. The most pronounced reduction in average power between baseline and the thermal stimuli was observed in the alpha band. Compared to baseline, the alpha power under the thermal stimuli had the greatest reduction in the left and right parietal regions. There was no statistically significant difference in the change of average power in the low beta band. Significant increases in average power between baseline and the thermal stimuli were observed in the high beta band and gamma band in the global brain. However, there was no clear pattern or significant differences in the change of average power as the temperature of the stimuli increased.

Table 1

The F and p values of the one-way repeated-measures ANOVA on the average power of the six frequency bands in the five brain regions under the five conditions: baseline and 45°C, 48°C, 51°C, and 54°C thermal stimuli
	Baseline vs. 45°C vs. 48°C vs. 51°C vs. 54°C
	Frontal	Central	Occipital	Left parietal	Right parietal
Delta	F_{4, 76}= 2.588, p = 0.043*	F_{4, 76} = 2.657, p = 0.039*	F_{4, 76} = 2.071, p = 0.093	F_{4, 76} = 1.708, p = 0.157	F_{2.8, 53.201} = 4.904, p = 0.005**
Theta	F_{4, 76} = 9.738, p < 0.001***	F_{4, 76} = 19.522, p < 0.001***	F_{2.346, 44.571} = 10.564, p < 0.001***	F_{4, 76} = 10.610, p < 0.001***	F_{2.888, 54.873} = 12.833, p < 0.001***
Alpha	F_{4, 76} = 3.235, p = 0.017*	F_{2.6, 49.392} = 9.232, p < 0.001***	F_{4, 76} = 3.379, p = 0.013*	F_{4, 76} = 10.260, p < 0.001***	F_{2.154, 40.919} = 17.951, p < 0.001***
Low beta	F_{4, 76} = 1.312, p = 0.273	F_{4, 76} = 1.564, p = 0.192	F_{3.043, 57.810} = 1.501, p = 0.210	F_{4, 76} = 1.961, p = 0.109	F_{4, 76} = 1.650, p = 0.0.171
High beta	F_{4, 76} = 3.007, p = 0.023*	F_{4, 76} = 2.564, p = 0.045*	F_{4, 76} = 3.900, p = 0.006**	F_{4, 76} = 3.417, p = 0.013*	F_{4, 76} = 3.395, p = 0.013*
Gamma	F_{4, 76} = 5.440, p < 0.001***	F_{4, 76} = 5.548, p < 0.001***	F_{4, 76} = 7.274, p < 0.001***	F_{4, 76} = 6.433, p < 0.001***	F_{4, 76} = 8.388, p < 0.001***

The above results for mean power indicate that significant changes in the EEG response to thermal stimuli occurred mainly in the theta and alpha frequency bands, and the most significant changes in the average power of the alpha band were observed at the highest temperature of the stimuli (54°C). This may indicate thermal stimulation had the greatest effect on and modulation of the power of the EEG alpha band. The results of the post-hoc comparisons showed there was a significant difference in the average power of the alpha band at stimulus temperatures of 48°C and 54°C. Therefore, the difference between the average alpha-band power of the thermal stimuli at 48°C and 54°C were compared to the baseline. Figure 3 shows a brain topographical map of the distribution of the mean power changes in the alpha band. Many studies using the pain stimulation paradigm have shown that the primary somatosensory cortex (SI) and secondary somatosensory cortex (SII) consistently respond to pain [19, 20]. It was obvious that the reduction in alpha-band power was mainly concentrated in the contralateral somatosensory area near electrodes C3 and C4 (SI, SII). The results indicate that the greatest amount of change in mean alpha-band power in response to thermal stimulation occurred in these two brain regions at 48°C and 54°C.

3.2 Results of Granger Causality Analysis

The $DTF$ matrices were calculated to explore the differences between temperature and the strength of the causal connections at each node of the network. After the average $DTF$ connectivity matrix of EEG data from healthy subjects under the five conditions (baseline and 45°C, 48°C, 51°C, and 54°C thermal stimuli) were obtained, the parameter $DTF$_mean values characterizing the strength of causal connections in the global brain were calculated.

Figure 4 shows the $DTF$_mean values of the multichannel EEGs in each frequency band (the delta, theta, alpha, low beta, high beta, and gamma bands) under the five conditions (baseline and 45°C, 48°C, 51°C, and 54°C thermal stimuli). Table 2 shows the results of the one-way repeated-measures ANOVA and Table 3 shows the post-hoc comparisons for the $DTF$_mean values of the six frequency bands under the five conditions. In all six frequency bands, there were significant differences in the $DTF$_mean values of baseline, 45°C, 48°C, 51°C, and 54°C thermal stimuli. The results of the post-hoc comparisons showed there was a significant reduction in $DTF$_mean values in the delta, theta, alpha and gamma bands as the temperature of the stimuli increased (except for 45°C). Significant differences in the $DTF$_mean values between the thermal stimuli were found at 48°C and 54°C (delta, p = 0.044; alpha, p = 0.007; low beta, p = 0.005; and gamma, p = 0.008). This suggests that a thermal stimulus leads to reduced network connectivity in healthy subjects. In the low beta and high beta bands, there was no significant difference in $DTF$_mean values under thermal stimulation compared to baseline. Moreover, there was no significant trend in $DTF$_mean values with increasing stimulus temperature. Since the $DTF$_mean values had the most significant differences in the alpha band, clustering coefficients and the global efficiency of the DTF causal network of the alpha band for the five conditions were calculated and compared (Fig. 5). Table 4 shows the results of the one-way repeated-measures ANOVA and the post-hoc comparisons for the clustering coefficient and the global efficiency for baseline, 45°C, 48°C, 51°C, and 54°C stimuli. In the alpha band, the clustering coefficient and global efficiency were significantly smaller for the highest temperature stimulus than those for baseline and lower temperature stimuli. These results suggest that, compared to baseline, the brain networks of healthy subjects receiving thermal stimuli manifest sparser connectivity and lower global information transfer efficiency in the alpha band. Compared to higher temperature stimuli, the brain networks of healthy subjects receiving lower temperature stimuli manifest denser connectivity and higher global information transfer efficiency.

Table 2

The F and p values of the one-way repeated-measures ANOVA and p values of the post-hoc tests for the $DTF$_mean values of the six frequency bands under the five conditions of baseline, 45°C, 48°C, 51°C, and 54°C thermal stimuli
	$DTF$_mean Baseline vs. 45°C vs. 48°C vs. 51°C vs. 54°C
	One-way repeated-measures ANOVA
Delta	F_{4, 76} = 6.071, p < 0.001***
Theta	F_{4, 76} = 5.109, p = 0.001**
Alpha	F_{2.798, 53.167} = 10.067, p < 0.001***
Low beta	F_{4, 76} = 4.033, p = 0.005**
High beta	F_{4, 76} = 2.557, p = 0.045*
Gamma	F_{4, 76} = 9.663, p < 0.001***

Table 3

The p values of the post-hoc tests for the $DTF$_mean values of the six frequency bands under the five conditions of baseline, 45°C ,48°C, 51°C and 54°C thermal stimuli
p value	Post-hoc test
	B-45	B-48	B-51	B-54	45–48	45–51	45–54	48–51	48–54	51–54
Delta	0.008**	1.000	0.02*	0.003**	0.647	1.000	1.000	1.000	0.044*-	1.000
Theta	0.009**	0.346	0.036*	0.005**	1.000	1.000	1.000	1.000	0.084	1.000
Alpha	0.007**	0.005**	0.010*	0.001***	1.000	1.000	0.969	1.000	0.007**	0.363
Low beta	0.633	1.000	1.000	0.041*	0.195	1.000	1.000	0.150	0.005*	1.000
High beta	1.000	1.000	1.000	0.154	0.437	1.000	1.000	0.905	0.103	1.000
Gamma	0.005**	0.059	0.001**	0.001***	0.769	1.000	1.000	0.239	0.008**	1.000

3.3 the VAS Results

Table 5 shows the results of the subjective pain ratings on the VAS scale. The result showed that the subjective pain ratings of healthy subjects increased significantly with increasing stimulus temperature. Most subjects reported their pain perception of the 50° C stimulus was maximal during the first one or two seconds of stimulation. After that, their perception of pain gradually decreased as they adapted to it.

Table 4

The F and p values of the one-way repeated-measures ANOVA and p values of the post-hoc tests for the clustering coefficient and the global efficiency of baseline, 48°C, and 54°C stimuli
Baseline vs. 45°C vs. 48°C vs. 51°C vs. 54°C
		Clustering coefficient	Global efficiency
One-way repeated-measures ANOVA		F_{2.450, 46.542} = 8.602, p < 0.001***	F_{2.722, 51.710} = 7.314, p < 0.001***
Post-hoc test	B-45	P = 1.000	P = 0.144
	B-48	P = 0.194	P = 0.019*
	B-51	P = 0.074	P = 0.003**
	B-54	P < 0.001***	P < 0.001***
	45–48	P = 0.318	P = 1.000
	45–51	P = 0.159	P = 1.000
	45–54	P = 0.008**	P = 0.872
	48–51	P = 1.000	P = 0.531
	48–54	P = 0.006**	P = 0.060
	51–54	P = 0.647	P = 1.000

Table 5

Differences between 45°C, 48°C, 51°C, and 54°C on the VAS
	45°C Mean (SD)	48°C Mean (SD)	51°C Mean (SD)	54°C Mean (SD)	Significances
VAS	0.63 (0.72)	3.83 (2.06)	5.77 (1.44)	7.1 (0.91)	F_{1.676, 23.471} = 105.628, p < 0.001***

This experiment was designed to determine the effect of different degrees of thermal stimulation applied to the oral mucosa on EEG in healthy human subjects. Changes in the EEG spectrum were examined in different brain regions by frequency domain analysis. The results showed that the EEG responses of healthy subjects to thermal stimulation primarily occurred in the theta and alpha bands. The average power change in the alpha band was most pronounced at the highest stimulus temperature. It was obvious that the reduction of alpha power was mainly concentrated in the SI and SII, near electrodes C3 and C4. In addition, we used DTF-based Granger causality analysis to explore the connectivity of brain networks induced by stimuli of different temperatures. The results of that analysis showed that high temperature stimuli reduced network connectivity and global information transfer efficiency.

Numerous studies have been conducted to evaluate the effect of tonic heat pain on human body parts, including the fingers [21, 22], forearm [1], the dorsum of the hands [23, 24], and the thigh [25, 26]. In most cases, they were mainly concerned about skin irritation. The oral mucosa is the main barrier against oral diseases, and it is worth studying how the EEG signals from the CNS respond to high temperature conditions. Sustained thermal stimulation was applied to the oral mucosal in this study. The EEG findings are similar to those of studies of transient stimuli (< 40 ms), which have reported a decrease in alpha power [27, 28]. A previous study investigated the local inhibition of alpha frequency oscillatory brain activity in the sensorimotor cortex during tonic pain [23]. Similar to that study, our study found a reduction in alpha power, but we also found a reduction in theta power and a local increase in gamma power. Schulz et al. found the gamma rhythm to be positively related to subjective pain ratings [29]. In addition to analyzing thermal stimuli from a power perspective, we investigated changes in network connectivity and transmission efficiency caused by thermal pain stimuli and used Granger causality analysis to study the relationship between thermal stimulation and brain networks from the direction of information flow in neural networks.

In previous studies, thermal pain stimuli were always accompanied by power changes in different frequency bands. Most studies reported increased delta power during pain using thermal stimuli [1, 25, 30–32]. However, our study did not find a significant change in delta power. In other studies, thermal stimulation was mainly applied to the outer epidermis, such as the hands or legs, whereas our stimulation was applied to the oral mucosa. We assumed that thermal stimulation applied to the skin and oral mucosa is different because the tissue structure of the skin and oral mucosa is different, and the temperature stimulation is transmitted to the CNS differently, which leads to different proportions of EEG frequency components. Huber et al. [1] and Zhang et al. [33] both found reduced theta activity to tonic heat pain in the hand, which is consistent with our results. Most EEG studies related to tonic heat pain have found decreased alpha power [1, 33–35]. There is some research evidence that the alpha band has a functional inhibitory effect; i.e., the alpha band activity gates information by inhibiting task-irrelevant regions of the brain [36, 37]. This is consistent with the present findings, and we also found suppression of small-world networks in the alpha band, with reduced connectivity and transmission efficiency. Alpha waves often occur when the body is in a state of quietness and relaxation. Some research indicates that the alpha band plays a key role in invoking responses in higher hierarchical nodes of the brain, such as the anterior cingulate cortex (ACC), which is associated with the emotional and perceptual aspects of pain processing [33, 38]. We assumed that when thermal stimulation is applied, subjects perceive the thermal stimulation, start to feel uncomfortable, even nervous. Our study showed a significant difference between 48°C and 54°C thermal stimulation, regardless of the power spectrum or the small-world network. This leads us to speculate that the CNS response to thermal stimulation of the oral mucosa requires a large temperature difference to achieve. Subsequently, we will refine the temperature gradient more to study the EEG response to oral thermal stimulation at specific temperatures.

Most research during the last decades has revealed an extended network of brain areas related to different aspects of pain [20, 39]. This network essentially consists of the primary (S1) and secondary (S2) somatosensory cortices, the insular cortex, the ACC, and prefrontal areas. We found small-world network connectivity degradation occurred in low medium frequency bands in the presence of a painful stimulus. Zhang et al. reported a consistent information flow initiated from the contralateral sensorimotor region to the medial frontal area in the presence of a painful stimulus; in contrast, when there was no pain, such a connectivity pattern was absent [11]. The temperature receptors in the cerebral cortex are located in the paracentral part of the parietal cortex of the posterior central gyrus, and these areas are referred to as the somatosensory region I [40]. The present research found a strong correlation between SI, SII, and pain, which is consistent with the consensus that SI and SII are part of the pain network.

Many studies related to pain stimulation networks have stimulated the external epidermis of subjects, such as the palms, back of the hands, thighs, or the whole body. Few studies have reported pain networks associated with oral thermal stimulation [5, 41]. The central regions of the maxillary and mandibular incisors, which are the most temperature-sensitive areas of the oral mucosa were selected to be the stimulation areas in this study. The oral mucosa, which was the site of stimulation, is innervated by the trigeminal nerve [3, 4]. Physiological studies have confirmed that the trigeminal nerve provides afferent information from the skin, sinuses, mucous membranes, and oral cavity about temperature, touch, and pain [42]. The oral cavity receives intraoral temperature and cold sensory stimuli through mucosal receptors, which converge on the caudate nucleus of the bilateral trigeminal tract through tympanic, trigeminal, and lingual nerve fibers. Then, the caudate nucleus of the spinal trigeminal nucleus further transmits warm pain information to the ventromedial and ventropostolateral nuclei of the thalamus, which transmits the information to the cerebral cortex after neuronal changes [3, 4]. A study by Li et al. reported that the oral thermal response of parabrachial taste neurons could be reversibly suppressed by optogenetic-assisted inhibition of the trigeminal neural circuit, implying that thermal related neural activity has a trigeminal origin [43]. The present study, which designed an experimental paradigm for applying thermal stimulation to the oral mucosa, could provide a theoretical basis and technical support for studying the mechanism of transmission and perception of oral thermal pain stimulation through a trigeminal pathway to the CNS in human subjects. BMS is a pathological pain, belongs to chronic pain syndrome, peripheral and central nervous system plasticity. Neuropathic pain in the CNS is one of the mechanisms of BMS [44]. Some studies have suggested that the burning sensation in patients with BMS is due to malfunctioning neural networks that produce pain experiences in the brain [45, 46]. Our study was conducted to observe changes in the central system under thermal pain stimulation of the oral cavity and resulted in structural network changes in contralateral somatosensory areas, which are consistent with neural network changes in BMS.

Limitations

The present research was conducted only on college students, and the subjects were not distributed across all age groups. Moreover, the subjects were healthy, and subjects with oral disease-related conditions need to be recruited for comparison with healthy subjects in the follow-up studies. The cold and hot temperature perception thresholds show a gradual increase from the midline to both sides of the mouth, and the left and right sides are symmetrical [5]. However, because it is too complicated to compare different temperature stimuli in different subdivisions of the oral mucosa, we chose only to use the central areas of the maxillary and mandibular incisor areas, which are the most sensitive to thermal stimuli. We focused on the relationship between changes in alpha power and painful stimuli because the alpha band has been widely reported in the literature to be associated with pain networks [1, 11, 21, 34, 35]. However, changes in higher frequency bands were found in our study, which also deserve to be examined in subsequent studies. The present study only observed the strength of whole-brain causal connections, and did not obtain ideal results for the information flow of nodes in the network. Therefore, a more detailed investigation of various information in the pain network is expected to be conducted in follow-up studies.

In conclusion, the current study of thermal stimulation of the oral mucosa in healthy subjects showed that changes in EEG power were concentrated in the theta and alpha bands. The alpha power decreased the most in the contralateral somatosensory area. Furthermore, the results of Granger causal analysis showed that the connectivity and global transmission efficiency of the brain's small-world network were negatively correlated with pain levels. This provides a foundation for studying the distribution and rhythmic EEG activity related to thermal pain stimuli in brain regions, as well as the properties of brain network connections associated with pain, and should help broaden understanding of cortical responses to thermal pain stimuli, from an EEG signal perspective.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 61977050).

Ethical Approval

This study was approved by the ethics committee of Tianjin Medical University (No. TMUhMEC20210201) and was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki.

Consent to participate

Informed consent was obtained from all individual participants included in the study.

Consent to publish

Patients signed informed consent regarding publishing their data and photographs.

Conflict of Interest

The authors declare no competing interests.

Authors' contributions

SJ. H. was responsible for study conception and design, device design and assemble, H.Z. was responsible for volunteer recruitment, data analyses, and the drafted manuscript; G. C. contributed to the data interpretation and data analyses; YX. H. contributed to the critical revision of the manuscript; SG. W. contributed to the study conception and design, data interpretation, critically reviewing manuscript and crucial guidance in the study. All authors have given final approval and agreed to be responsible for all aspects of this study to ensure accuracy and integrity.

Funding

This study was funded by the National Natural Science Foundation of China (No. 61977050).

Availability of data and materials

Not applicable

Huber, M.T., et al., EEG responses to tonic heat pain. 2006. 173(1): p. 14–24.
Yamada, M., J. Maruhashi, and N. Miyake, The distribution of sensory spots on the oral mucous membrane. Jpn J Physiol, 1952. 2(4): p. 328–32.
Darian-Smith, I., et al., Warm fibers innervating palmar and digital skin of the monkey: responses to thermal stimuli. J Neurophysiol, 1979. 42(5): p. 1297–315.
Kawahara, K., Y. Sawada, and M.J.B.R. Aoki, Dual pathways for thermal afferents from the cat's tongue. 1986. 378(1): p. 61–68.
Wang, Z.G., et al., [Correlation of perceptive temperature threshold of oral mucosa and sympathetic skin response]. Zhonghua Kou Qiang Yi Xue Za Zhi, 2018. 53(2): p. 79–84.
The International Classification of Headache Disorders, 3rd edition (beta version). Cephalalgia, 2013. 33(9): p. 629–808.
T, S., et al., Spatial and Temporal Brain Responses to Noxious Heat Thermal Stimuli in Burning Mouth Syndrome. 2016. 95(10): p. 1138–46.
Mulders, D., et al., Dynamics of the perception and EEG signals triggered by tonic warm and cool stimulation. PLoS One, 2020. 15(4): p. e0231698.
Peng, W., et al., Changes of spontaneous oscillatory activity to tonic heat pain. PLoS One, 2014. 9(3): p. e91052.
Akin, M. and M.K. Kiymik, Application of Periodogram and AR Spectral Analysis to EEG Signals. 2000. 24(4): p. 247–256.
Huishi Zhang, C., et al., Spectral and spatial changes of brain rhythmic activity in response to the sustained thermal pain stimulation. Hum Brain Mapp, 2016. 37(8): p. 2976–91.
Kaminski, M.J. and K.J.J.B.C. Blinowska, A new method of the description of the information flow in the brain structures. 1991. 65(3): p. 203–210.
Bassett, D.S. and E.J.T.N. Bullmore, Small-World Brain Networks. 2007. 12(6): p. 512–523.
Wilke, C., et al., Neocortical seizure foci localization by means of a directed transfer function method. 2010. 51(4): p. 564–572.
He, B., et al., eConnectome: A MATLAB toolbox for mapping and imaging of brain functional connectivity. 2011. 195(2): p. 261–269.
Xu, X., et al., Inhibition of propofol anesthesia on functional connectivity between LFPs in PFC during rat working memory task. PLoS One, 2013. 8(12): p. e83653.
Barrat, A., et al., The architecture of complex weighted networks. 2004. 101(11): p. 3747–3752.
Latora, V. and M.J.P.R.L. Marchiori, Efficient Behavior of Small-World Networks. 2001. 87(19): p. 198701.
Wager, T.D., et al., An fMRI-based neurologic signature of physical pain. N Engl J Med, 2013. 368(15): p. 1388–97.
Tracey, I. and P.W. Mantyh, The cerebral signature for pain perception and its modulation. Neuron, 2007. 55(3): p. 377–91.
Lv, B., et al., Effects of stimulus mode and ambient temperature on cerebral responses to local thermal stimulation: An EEG study. Int J Psychophysiol, 2017. 113: p. 17–22.
An, K.M., et al., Magnetoencephalographic study of event-related fields and cortical oscillatory changes during cutaneous warmth processing. Hum Brain Mapp, 2018. 39(5): p. 1972–1981.
Nickel, M.M., et al., Neural oscillations and connectivity characterizing the state of tonic experimental pain in humans. Hum Brain Mapp, 2020. 41(1): p. 17–29.
Fauchon, C., et al., The Modular Organization of Pain Brain Networks: An fMRI Graph Analysis Informed by Intracranial EEG. Cereb Cortex Commun, 2020. 1(1): p. tgaa088.
Giehl, J., et al., Responses to tonic heat pain in the ongoing EEG under conditions of controlled attention. Somatosens Mot Res, 2014. 31(1): p. 40–8.
Bunk, S.F., et al., Does EEG activity during painful stimulation mirror more closely the noxious stimulus intensity or the subjective pain sensation? Somatosens Mot Res, 2018. 35(3–4): p. 192–198.
Hu, L., et al., Functional features of nociceptive-induced suppression of alpha band electroencephalographic oscillations. J Pain, 2013. 14(1): p. 89–99.
Ploner, M., et al., Oscillatory activity reflects the excitability of the human somatosensory system. Neuroimage, 2006. 32(3): p. 1231–6.
Schulz, E., et al., Prefrontal Gamma Oscillations Encode Tonic Pain in Humans. Cereb Cortex, 2015. 25(11): p. 4407–14.
Ferracuti, S., et al., Quantitative EEG modifications during the Cold Water Pressor Test: hemispheric and hand differences. Int J Psychophysiol, 1994. 17(3): p. 261–8.
Chang, P.F., L. Arendt-Nielsen, and A.C. Chen, Dynamic changes and spatial correlation of EEG activities during cold pressor test in man. Brain Res Bull, 2002. 57(5): p. 667–75.
Stevens, A., et al., Both pain and EEG response to cold pressor stimulation occurs faster in fibromyalgia patients than in control subjects. Psychiatry Res, 2000. 97(2–3): p. 237–47.
Zhang, H., et al., Spectral and spatial changes of brain rhythmic activity in response to the sustained thermal pain stimulation. Hum Brain Mapp, 2016. 37(8): p. 2976–91.
Navid, M.S., et al., The effects of chiropractic spinal manipulation on central processing of tonic pain - a pilot study using standardized low-resolution brain electromagnetic tomography (sLORETA). Sci Rep, 2019. 9(1): p. 6925.
Hansen, T.M., et al., Characterization of cortical source generators based on electroencephalography during tonic pain. J Pain Res, 2017. 10: p. 1401–1409.
Klimesch, W., P. Sauseng, and S. Hanslmayr, EEG alpha oscillations: the inhibition-timing hypothesis. Brain Res Rev, 2007. 53(1): p. 63–88.
Jensen, O. and A. Mazaheri, Shaping functional architecture by oscillatory alpha activity: gating by inhibition. Front Hum Neurosci, 2010. 4: p. 186.
Giehl, J., et al., Responses to tonic heat pain in the ongoing EEG under conditions of controlled attention. 2014. 31(1): p. 40–48.
Craig, A.D., Pain mechanisms: labeled lines versus convergence in central processing. Annu Rev Neurosci, 2003. 26: p. 1–30.
Brownsett, S. and R.J.C.C. Wise, The Contribution of the Parietal Lobes to Speaking and Writing. 2010. 20(3): p. 517–523.
Lemon, C.H., Tasting temperature: neural and behavioral responses to thermal stimulation of oral mucosa. Curr Opin Physiol, 2021. 20: p. 16–22.
Klein, A.H., The orotrigeminal system. Handb Clin Neurol, 2019. 164: p. 205–216.
Li, J. and C.H. Lemon, Mouse Parabrachial Neurons Signal a Relationship between Bitter Taste and Nociceptive Stimuli. J Neurosci, 2019. 39(9): p. 1631–1648.
Wada, A., et al., Altered structural connectivity of pain-related brain network in burning mouth syndrome-investigation by graph analysis of probabilistic tractography. Neuroradiology, 2017. 59(5): p. 525–532.
Kohashi, R., et al., Time-dependent responses in brain activity to ongoing hot stimulation in burning mouth syndrome. J Oral Sci, 2020. 62(2): p. 170–174.
Yoshino, A., et al., Functional Alterations of Postcentral Gyrus Modulated by Angry Facial Expressions during Intraoral Tactile Stimuli in Patients with Burning Mouth Syndrome: A Functional Magnetic Resonance Imaging Study. Front Psychiatry, 2017. 8: p. 224.

No competing interests reported.

SupplementaryMaterial.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

A study on the effects of thermal pain stimulation at the oral mucosa in EEG

Status:

Version 1

Abstract

Objectives

Materials and methods

Results

Conclusions

Clinical Relevance

Figures

1 Introduction

2 Materials And Methods

2.1 Participants

2.2 the Instrumentation of the Experiment

2.3 Experimental Paradigm

2.4 Data Analysis

2.5 Statistical Analysis

3 Results

3.1 Results of the Frequency Analysis

3.2 Results of Granger Causality Analysis

3.3 the VAS Results

4 Discussion

5 Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1