Objective The aim of the present work was to observe the activation/deactivation of cerebral functional regions after electroacupuncture (EA) at Yintang (EX-HN3) and GV20 by functional MRI (fMRI).
Design A total of 12 healthy volunteers were stimulated by EA at Yintang and GV20 for 30 min. Resting-state fMRI scans were performed before EA, and at 5 and 15 min after needle removal. Statistical parametric mapping was used to preprocess initial data, and regional homogeneity (ReHo) and amplitude of low-frequency fluctuation (ALFF) were analysed.
Results ReHo at 5 min post stimulation showed increases in the left temporal lobe and cerebellum and decreases in the left parietal lobe, occipital lobe and right precuneus. At 15 min post stimulation, ReHo showed increases in the left fusiform gyrus; lingual gyrus; middle temporal gyrus; postcentral gyrus; limbic lobe; cingulate gyrus; paracentral lobule; cerebellum, posterior lobe, declive; right cuneus and cerebellum, anterior lobe, culmen. It also showed decreases in the left frontal lobe, parietal lobe, right temporal lobe, frontal lobe, parietal lobe and right cingulate gyrus. ALFF at 5 min post stimulation showed increases in the right temporal lobe, but decreases in the right limbic lobe and posterior cingulate gyrus. At 15 min post stimulation ALFF showed increases in the left frontal lobe, parietal lobe, occipital lobe, right temporal lobe, parietal lobe, occipital lobe and cerebellum, but decreases in the left frontal lobe, anterior cingulate gyrus, right frontal lobe and posterior cingulate gyrus.
Conclusions After EA stimulation at Yintang and GV20, which are associated with psychiatric disorder treatments, changes were localised in the frontal lobe, cingulate gyrus and cerebellum. Changes were higher in number and intensity at 15 min than at 5 min after needle removal, demonstrating lasting and strong after-effects of EA on cerebral functional regions.
- Statistics & Research Methods
Statistics from Altmetric.com
Acupuncture, a vital component of traditional Chinese medicine (TCM), has become an indispensable part of modern medicine because its treatment effects are increasingly being recognised. However, the poor elucidation of acupuncture mechanisms remains an impediment to its further acceptance. Thus, advances in the scientific knowledge of the mechanisms of acupuncture treatment are urgently needed.
With the development of functional MRI (fMRI), changes in blood flow in cerebral functional regions can be monitored by the blood oxygenation-level-dependent (BOLD) method. Consequently, the central mechanisms of acupuncture treatment can be analysed using this method.
Block design is a frequently used method in fMRI research. Recent reports have adopted the block design, in which fMRI head scanning is carried out comparing the responses to alternating stimulation and release, which we know as ‘manipulating the needle–retaining the needle’.1–4 Block design has led to numerous significant discoveries and conclusions and has confirmed the interrelationships between acupuncture, acupuncture points and the brain. However, this design ignores the effects of acupuncture on functional changes in cerebral regions while the needles are being retained. Comparison between states of stimulating and retaining needles examines only the effects of stimulation, rather than the overall effect of clinical acupuncture in which the needles are generally manipulated and retained. After stimulating the needle, cerebral activities do not return to their previous stable state, but remain unstable while the needles are being retained. This may lead to false conclusions about stimulation effects because it the effects of needle retention are not considered.
Resting-state fMRI experimental design is a new method that is gradually being applied in acupuncture research.5 ,6 In resting-state fMRI, subjects undergo scanning while they are conscious and recumbent, with the entire body relaxed, without being administered a cognitive task and without any systematic thinking. Unlike the traditional ‘acupuncture-response’ signal-processing mode, resting-state fMRI design manifests signals different from the linear hypothesis. These signals are closest to true cerebral responses under an analogous physiological state and reflect complex cerebral activities. Therefore, we adopted a resting-state fMRI design, rather than a block design, with the abovementioned alternating mode to directly observe the actual blood flow under specific conditions, rather than comparing stimulating and baseline states.
One essential factor for the treatment effects of acupuncture is the stimulation technique. In research, manual acupuncture is commonly used because cerebral imaging acupuncture studies primarily employ the block design, which involves manipulating needles. There are reports indicating that activations in cerebral regions differ between electroacupuncture (EA) and manual acupuncture.7 Considering the expanding clinical applications of EA and the scarcity of cerebral functional imaging research on it, we adopted EA for our research.
Currently, cerebral functional imaging technology is primarily used for research on the distinctive features of acupuncture points, including observations of the variations in post-stimulation effects on cerebral functional regions between acupuncture points and non-acupuncture points and between different acupuncture points.8–12 Most of this research has concentrated on cerebral functional imaging with single point acupuncture, whereas in clinical practice, acupuncture treatment usually involves a combination of acupuncture points. Because two acupuncture points is the minimum possible combination, we focused on the effect of the acupuncture point combination Yintang (EX-HN3) and Baihui (GV20).
The combination of Yintang and GV20 is frequently used in clinical practice and is especially effective in treating psychiatric disorders. According to TCM, GV20 is the acupuncture point of the GV meridian, which is ‘good for the brain and spirit’. Yintang is the acupuncture point of the Extra meridian, which ‘can calm and enlighten the mind’. The literature and recent reports indicate that Yintang and GV20 usually serve as the main acupuncture points in the treatment (as adjunctive therapy) of psychiatric disorders, with confirmed treatment effects.12–15 Our experiments with resting-state fMRI design complement other fMRI studies on this acupuncture point combination, because block design experiments cannot incorporate real-time fMRI scanning of the cerebral regions because these two acupuncture points are located on the head.
To analyse imaging data, we have adopted regional homogeneity (ReHo) and amplitude of low-frequency fluctuation (ALFF) data analysis methods, which can analyse the initial signals of the whole brain from different perspectives. ReHo indicates the temporal similarity between a voxel and its neighbouring voxels and reflects the coordinating function of the idiopathic activity of the cerebral neurons from the perspective of activity homogeneity.16 ALFF can directly demonstrate blood oxygenation levels and reflects the idiopathic activity levels of the neurons in the voxels according to their energy under the resting state.16 ,17 Therefore, combining these two methods to analyse the effects of acupuncture on cerebral functional regions should lead to a more comprehensive conclusion.
Thus, in our study, after healthy volunteers were administered EA stimuli at Yintang and GV20 for 30 min, resting-state fMRI scans were obtained before stimulation, and 5 and 15 min after needle removal to observe changes in cerebral functional regions. ReHo and ALFF analyses were performed to explore the central mechanism of the EA effects of Yintang and GV20.
Materials and methods
The 12 healthy volunteers, aged 20–24 years (6 men), were all university students. All were right handed, of normal body weight and height and without addictions to coffee, cigarettes or alcohol. No subject had received acupuncture treatment within 4 weeks of the experiment. The subjects underwent an acupuncture response test, in which they were stimulated at the same acupuncture points and the sensation scored by visual analogue scale (VAS), to exclude those with no response or an over-reaction to acupuncture. The subjects were informed about the experiment and voluntarily signed informed consent in advance. This experiment was approved by the Chinese Ethics Review Committee (ChiERCT-2011003) and registered at the Chinese Clinical Trial Registry (ChiCTR-TNRC-11001311).
Huatuo stainless steel acupuncture needles (25 mm×0.30 mm; Suzhou Acupuncture & Moxibustion Appliance Co. Ltd., Suzhou, China), Han's Acupoint and Nerve Stimulator (LH-202H; Beijing Hua Wei Industrial Development Corporation, Beijing, China), a 3.0-T MRI scanner (GE Signa Excite system, Connecticut, USA), specially designed earplug (Aearo Corporation, Indiana, USA) and eyeshades (Shinewon Travelling Products Factory, Hanjiang, Yangzhou City, China) were used in this study.
EA at Yintang and GV20
EA was applied at Yintang and GV20.18 At Yintang a needle was inserted tangentially to the skin, 0.5 cun towards the nose tip. At GV20, another needle was inserted tangentially to the skin 0.5 cun posteriorly towards GV19. Manual rotation was maintained in clockwise and counterclockwise directions until de qi was elicited. The needles were then connected to the electrodes of the stimulator. The stimulation frequency was fixed at 2/100 Hz with alternating dispersed and dense waves and current intensity was selected depending on the subject's tolerance (slight quivering of the skin). The EA continued for 30 min. The entire process (including the initial screening phase) was carried out by the same acupuncturist, who was an attending doctor.
Resting-state fMRI scanning
The subjects were awake, with normal respiration, and lay supine on an examination bed. The head was placed in a foam headrest for maximum restriction of passive and active movements of the head. The subjects were instructed to avoid any systematic mental activity, and visual and audio stimulations were minimised with earplugs and eyeshades. Scanning was initiated once the subjects were familiarised with the circumstances.
The fMRI statistics of cerebral function were gathered with the MRI scanner and an eight-channel head coil 30 min before stimulation, and 5 and 15 min after needle removal. Details of fMRI were as follows: T1W1 sequence for 3 min, TR 2300 ms (repetition time), TE 920 ms (echo time), field of vision (FOV) 240 mm×240 mm, thickness 5 mm, gap 1.0 mm and image matrix 320×256. The resting-state fMRI BOLD statistics were obtained by 6 min of scanning with a gradient-Echo Planar Imaging (GR-EPI) sequence, with parameters TR 3000 ms/TE 20 ms, flip angle (FA) 90°, FOV 240 mm×240 mm, layer thickness 5 mm, gap 1.0 mm and image matrix 96×96. Each scanning comprised approximately 24 layers. Anatomical image data were obtained by 6 min of scanning with a 3D fast spoiled gradient echo (FSPGR) sequence, with parameters TR 7.6 ms/TE 3.3 ms, FOV 240 mm×180 mm, FA 02°, FOV 256 mm×256 mm, layer thickness 1.2 mm and 248 layers in total.
Data analysis and statistical methods
The initial resting-state fMRI statistics were preprocessed with SPM8, including Digital Imaging and Communications in Medicine (DICOM) format conversion, layer timing, realignment, spatial standardisation and spatial smoothing. Realignment calculated the translation on the axes X, Y, Z and head motions during the scanning to exclude those with 3D translation of >1.5 mm or 3D rotation >1.5°. Spatial standardisation employed the Montreal Neurological Institute (MNI) tabulate developed by the MNI to observe the standardised fitting degree and exclude those with lower standardised fitting degrees. Spatial smoothing was carried out with the 4 mm×4 mm×4 mm Gaussian kernel. After data preprocessing, 10 cases were adopted for further statistical analysis, eliminating the 2 with lower standardised fitting degrees for severe head motion. (ReHo analysis was carried out after spatial standardisation without a spatial smoothing procedure).
After spatial standardisation, the linear trends were removed from the preprocessed data using the linear regression method with REST 1.6 software. The extracted time curve was then convolution processed with a Hamming band-pass filter to obtain the low-frequency oscillation signal amplitude (0.01–0.08 Hz). Subsequently, we calculated the Kendall's coefficient of concordance (KCC) value of the voxel to obtain the individual KCC map or ReHo map. These maps underwent whole-brain equalisation for further statistical analysis.16
After the spatial smoothing process, the linear trends were removed from the preprocessed data using the linear regression method with REST 1.6 software. The extracted time curve was then convolution processed with a Hamming band-pass filter to obtain the low-frequency oscillation signal amplitude (0.01–0.08 Hz). Subsequently, we calculated the ALFF value of the whole-brain voxels to obtain the ALFF map, which was divided by the average ALFF value. The standardised ALFF map was then delineated.17
Using SPM8 software, a two-sample t test was performed to investigate the standardised ReHo and ALFF differences between scans taken before stimulation, and 5 and 15 min after needle removal. Voxels with a combined threshold of p<0.005 and cluster size ≥10 voxels were considered significantly different. The ReHo and ALFF changes in the cerebral region were determined. Then, XjView 8.0 software was used to match the two-sample t test results to the statistically significant cerebral region on the MNI coordinate. Finally, the t test results were graphed using MRIcroN software.
ReHo at 5 min after needle removal
Compared with ReHo before EA, 5-min post-stimulation ReHo showed increases in the left superior temporal lobe and cerebellum, posterior lobe, declive. It showed decreases in the left inferior parietal lobule, middle occipital gyrus and the right parietal lobe precuneus (table 1, figure 1).
ReHo at 15 min after needle removal
Compared with ReHo before EA, 15-min post-stimulation ReHo showed increases in the left fusiform gyrus; lingual gyrus; middle temporal gyrus; postcentral gyrus, limbic lobe; cingulate gyrus; paracentral lobule; cerebellum, posterior lobe, declive; right cuneus and cerebellum anterior, lobe, culmen. It showed decreases in the left frontal lobe, superior frontal gyrus; parietal lobe, inferior parietal lobule; right temporal lobe, subgyral area; frontal lobe, subgyral area; posterior cingulate gyrus; superior frontal gyrus; middle frontal gyrus and parietal lobe, precuneus (table 2, figure 2).
ALFF at 5 min after needle removal
Compared with ALFF before EA, 5-min post-stimulation ALFF showed increases in the right superior temporal gyrus and decreases in the right limbic lobe and posterior cingulate gyrus (table 3, figure 3).
ALFF at 15 min after needle removal
Compared with ALFF before EA, 15-min post-stimulation ALFF showed increases in the left paracentral lobule; paracentral lobule; inferior parietal lobule; middle occipital gyrus; lingual gyrus; right fusiform gyrus; postcentral gyrus; parietal lobe, subgyral area; middle occipital gyrus and right cerebellum, posterior lobe, declive (table 4, figure 4).
Compared with ALFF before EA, 15-min post-stimulation ALFF showed decreases in the left superior frontal gyrus, middle frontal gyrus, the frontal lobe subgyral area, the anterior cingulate gyrus, the right middle frontal gyrus, the inferior frontal gyrus and the posterior cingulate gyrus (table 5, figure 4).
After EA at Yintang and GV20, ReHo showed that activations/deactivations were relatively centralised in Brodmann area19 (BA)2, 3, 4, 5, 6, 7, 8, 9, 10, 17, 18, 19, 21, 22, 23, 31, 37, 39, 40 and 46, along with the anterior and posterior lobes of the cerebellum. ALFF showed that activations/deactivations were relatively centralised in BA2, 3, 4, 9, 10, 11, 17, 18, 19, 22, 23, 30, 31, 32, 37, 40, 47 and the posterior lobe of the cerebellum.
ReHo and ALFF are data analysis methods in common use for measuring cerebral function. However, ReHo and ALFF use different perspectives to analyse the data, which leads to some differences in findings. Since the two analyses found some common changes of cerebral functional regions, both were adopted to reduce inaccuracies and to provide reliable and comprehensive conclusions. They agreed in showing changes in BA2, 3, 4, 9, 10, 17, 18, 19, 22, 23, 31, 37 and 40. BA2 and 3 are associated with sensory perception. BA4 is associated with intentional activities. BA9, 10, 23 and 31 are related to association and integration. BA17, 18, 19 and 37 are associated with visual information. BA22 and 40 are associated with vision and language reception. These preliminary results indicate that EA at Yintang and GV20 exerts specific effects in the cerebral regions related to perception, implementation, mental association, vision and hearing.
Duman has stated that the cortex of the frontal lobe is pertinent to emotional consciousness and experiences and is responsible for initiating, testing and modifying emotion, repressing emotional expression and integrating emotional perception.20 We have found that after EA at Yintang and GV20 in healthy people, ReHo and ALFF in the prefrontal lobe show decreasing trends, which may be one of the mechanisms by which EA at Yintang and GV20 has effects on psychiatric disorders. One study has shown that resting-state ReHo in the middle frontal gyrus, inferior frontal gyrus, left and right frontal lobes and lenticular nuclei of both sides are higher in patients with depression than in healthy people in the resting state,21 supporting our results.
As a part of the limbic system, the cingulate gyrus is a significant centre for emotional integration. It has a primarily granular cellular structure. The cingulate gyrus is involved in emotional control and awakening of perception, and is associated with self-monitoring and emotional regulation.22 ,23 We have found that the cingulate gyrus is activated after EA at Yintang and GV20, which may be related to the mechanism of treatment of psychiatric disorders by EA at these points, which may correspond with clinical therapeutic effects shown by Han et al.24
The posterior lobe of the cerebellum primarily regulates movement initiation, planning and coordination, including determining the power, direction and scope of movements. The cerebellum participates in emotional processing because of its significant role in initial discrimination of sensory and emotional information.25 ReHo and ALFF increase in the cerebellum region in healthy people following EA at Yintang and GV20, which could be related to another mechanism of treatment of psychiatric disorders by EA at Yintang and GV20. Other fMRI studies26 ,27 support the effect of acupuncture on the cerebellum.
In clinical practice, patients often receive acupuncture treatment once daily, suggesting that the therapeutic effects may last for a period after a single treatment. However, there have been no relevant studies demonstrating the existence of post-stimulation effects, the exact duration of such effects or their regulation. Our study shows that, compared with ReHo and ALFF before EA at Yintang and GV20, ReHo and ALFF increased in the cerebral region of healthy subjects 5 min after needle removal and significantly increased 15 min after needle removal. These preliminary results indicate the existence and maintenance of a post-stimulation effect, with a probable increasing trend from 5 to 15 min after needle removal. The activation/deactivation at 5 min are related to visual information, vision and language reception, association and integration, while the activation/deactivation at 15 min are mostly associated with sensory perception and intentional activities except regarding those areas mentioned above. This may correspond with Cordes et al's28 view that some of the relevant brain regions will form a closely related neural network after the stimulation effect of acupuncture. This could explain the differential brain response at 5 and 15 min. We plan to continue exploring this post-stimulation effect to determine its trends and peaks, as well as introducing the analytical method of repeated measures to analyse data from multiple time points.
There are limitations in our study. For example, we did not take the effect of true and sham needling into consideration, thus we cannot claim a point-specific effect. Our findings might relate to psychological factors, which will be excluded in further research. Secondly, the functional MRI of cerebral functional regions in healthy volunteers is likely to be different from that in patients with psychiatric disorders. Therefore, we will focus on post-stimulation effect of EA at Yintang and GV20 in patients with psychiatric disorders.
In conclusion, we have used ReHo and ALFF analyses to demonstrate the activation/deactivation of psychiatric-disorder-related cerebral functional regions, such as the frontal lobe, cingulate gyrus and cerebellum, using stimulation by EA at Yintang and GV20. We speculate that these results indicate possible mechanisms by which EA at these points brings about psychiatric and emotional adjustments.
The usual block design fMRI assesses only the effect of needle stimulation, not the overall process of needling.
Two new methods showed lasting effects of acupuncture on frontal cortex, cerebellum and cingulate.
The authors would like to express their gratitude to all those who have helped during the experiment and writing, especially Professor Yong Huang, Zhangjin Zhang and Director Yanping Chen at Nanfang Hospital, who directed and governed the fMRI scan in Nanfang Hospital. Enago (http://www.enago.cn) are thanked for the English language review.
Contributors YZ, SQ, JC, YH, ZZ: conception and design. LL, NW, GZ, XJ, SQ: acquisition of data or analysis and interpretation of data. YZ, SQ, LL, YH, NW: drafting the article or revising it critically for important intellectual content. YZ, SQ, YH: final approval of the version to be published.
Funding This work was supported by National 973 Program of China (grants no. 2006CB504505 and no. 2012GB518504) and supported by the Third Key Subject of the ‘211 Project’ of Guangdong Province, grant no. (2009)431.
Competing interests None.
Study approval This study was conducted with the approval of the Chinese Ethic Review Committee (ChiERCT-2011003) and is registered at the Chinese Clinical Trial Registry (ChiCTR-TNRC-11001311).
Ethics approval Chinese Ethic Review Committee (ChiERCT-2011003).
Patient consent Obtained.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.