ORIGINAL RESEARCH

published: 24 March 2022

doi: 10.3389/fneur.2022.827749

Transcutaneous Auricular Vagus

Nerve Stimulation Modulates the
Prefrontal Cortex in Chronic
Insomnia Patients: fMRI Study in the
First Session
Jia-Kai He 1,2† , Bao-Hui Jia 2† , Yu Wang 1 , Shao-Yuan Li 1 , Bin Zhao 1,3 , Zeng-Guang Zhou 4 ,
Yan-Zhi Bi 5 , Mo-Zheng Wu 1 , Liang Li 1 , Jin-Ling Zhang 1 , Ji-Liang Fang 6* and
Pei-Jing Rong 1*
1
Department of Physiology, Institute of Acupuncture and Moxibustion, China Academy of Chinese Medical Sciences, Beijing,
China, 2 Department of Acupuncture, China Academy of Chinese Medical Sciences Guang’anmen Hospital, Beijing, China,
Edited by: 3
Department of Acupuncture, Southern Medical University, Guangzhou, China, 4 Key Laboratory of Quantitative Remote
Jie Lu,
Sensing Information Technology, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, China,
Capital Medical University, China 5
Key Laboratory of Mental Health, Institute of Psychology, Chinese Academy of Sciences, Beijing, China, 6 Department of
Reviewed by: Radiology, China Academy of Chinese Medical Sciences Guang’anmen Hospital, Beijing, China
Hongbin Han,
Peking University Third Hospital, China
Baoci Shan, Objectives: Transcutaneous auricular vagus nerve stimulation (taVNS) has been
Institute of High Energy Physics reported to be effective for chronic insomnia (CI). However, the appropriate population
(CAS), China
for taVNS to treat insomnia is unclear.
*Correspondence:
Ji-Liang Fang Methods: Total twenty-four patients with CI and eighteen health controls (HC) were
[email protected]
recruited. Rest-state functional magnetic resonance imaging (Rs-fMRI) was performed
Pei-Jing Rong
[email protected] before and after 30 min’ taVNS at baseline. The activated and deactivated brain
† These
regions were revealed by different voxel-based analyses, then the seed-voxel functional
authors share first authorship
connectivity analysis was calculated. In the CI group, 30 min of taVNS were applied twice
Specialty section: daily for 4 weeks. Pittsburgh Sleep Quality Index (PSQI) and Flinders Fatigue Scale (FFS)
This article was submitted to were also assessed before and after 4 weeks of treatment in the CI group. The HC
Applied Neuroimaging,
a section of the journal group did not receive any treatment. The correlations were estimated between the clinical
Frontiers in Neurology scales’ score and the brain changes.
Received: 02 December 2021
Results: The scores of PSQI (p < 0.01) and FFS (p < 0.05) decreased after 4 weeks
Accepted: 28 February 2022
Published: 24 March 2022 in the CI group. Compared to the HC group, the first taVNS session up-regulated left
Citation: dorsolateral prefrontal cortex (dlPFC) and decreased the functional connectivity (FCs)
He J-K, Jia B-H, Wang Y, Li S-Y, between dlPFC and bilateral medial prefrontal cortex in the CI group. The CI groups’
Zhao B, Zhou Z-G, Bi Y-Z, Wu M-Z,
Li L, Zhang J-L, Fang J-L and
baseline voxel wised fMRI value in the dlPFC were negatively correlated to the PSQI and
Rong P-J (2022) Transcutaneous the FFS score after 4 weeks treatment.
Auricular Vagus Nerve Stimulation
Modulates the Prefrontal Cortex in Conclusions: It manifests that taVNS has a modulatory effect on the prefrontal cortex
Chronic Insomnia Patients: fMRI Study in patients with CI. The initial state of dlPFC may predict the efficacy for taVNS on CI.
in the First Session.
Front. Neurol. 13:827749. Keywords: chronic insomnia, transcutaneous auricular vagus nerve stimulation, functional magnetic resonance
doi: 10.3389/fneur.2022.827749 imaging (fMRI), biomarkers, prefrontal cortex, neuromodulation

Frontiers in Neurology | www.frontiersin.org 1 March 2022 | Volume 13 | Article 827749

He et al. taVNS Modulates the PFC in CI Patients

INTRODUCTION TABLE 1 | Sample characteristics of the participants.

Chronic insomnia (CI) disorder is categorized as primary or Items CI (N = 20) HC (N = 28) Z/χ 2 p-value

secondary, depending on whether the sleep problem is caused by

Age (year) 42.50 ± 15.42 43.5 ± 11.23 −0.278 0.781
another medical and mental disorder or medication substance
Sex (M/F) 8/12 6/12 0.181 0.671
use (1). The main treatments for CI are medications and
Education (year) 12.20 ± 4.62 12.83 ± 6.24 −0.179 0.858
physical therapies. Cognitive behavioral therapy (CBT), one
of the most mainstream physical therapy for insomnia, was Z, Wilcoxon rank testing; χ 2 , chi-square testing. CI, chronic insomnia; HC, healthy control.
found to reduce the Functional Connectivity (FC) between the
Ventral Medial Prefrontal Cortex (vmPFC) and the striatum
in patients suffering from insomnia (2). Medication also affects Our previous studies revealed that taVNS adjusts the frontal
brain activity. In healthy participants, zolpidem reduced the cortex, insular, PCC, and amygdala in patients with major
neural activity in occipital lobe during visual stimulation (3). depression disorder (25–28). The modulated brain regions were
Agomelatine and mirtazapine increased the FC between right also closely related to sleep. According to the hyperarousal
Dorsolateral Prefrontal Cortex (dlPFC) and right Precuneus in theory, patients with insomnia have an overexcited but low
Major Depression Disorder (MDD) patients with sleep disorder functioning cortex (18, 29), which leads to nocturnal sleep
(4). Physical therapies have fewer side effects and therapy disturbances, daytime fatigue, and low work efficiency (29). In
dependence. Guidelines of sleep disorder recommend physical this study, the instant effects of taVNS would be explored. We
therapies as the first treatment before medications (5–8). hypothesize that taVNS would modulate the forebrain, especially
Insomnia is also a risk factor for depression (9). It is often brain regions related to emotion and cognition in patients
accompanied by mental problems (6). Colleges have to pay close with CI.
attention to some potential curative effect of neuromodulations
on insomnia (10), which have been widely used in the treatment
of mental diseases. Deep Brain Stimulation (DBS) and Vagus
MATERIALS AND METHODS
Nerve Stimulation (VNS) are invasive neuromodulations, DBS Recruitment of Participants
was reported to have occasionally improved a patient’s sleep A total of twenty- four patients with CI were recruited. They
problems in a patient with Parkinson’s disease (11). Stimulating were diagnosed according to the Fifth Edition of the Diagnostic
the cat’s Nucleus Tractus Solitaries (NTS), the nucleus into and Statistical Manual of Mental Disorders (DSM-V, 2015). All
which sensory fibers of the vagus nerve mainly project, increases participants were right-handed. Before the study, they were all
the theta and beta band power of left amygdala and pre- informed of the study protocol and volunteered to participate in
frontal cortices. As a result, the cats performed an increase in the study. Patients with fMRI contraindications, severe organic
wakefulness and a total time of rapid eye movement (REM) or mental diseases were excluded. Patients would voluntarily quit
sleep (12). These suggest that DBS and VNS have potentially the ongoing therapies including sleeping pills for at least 2 weeks.
curative effect on insomnia. Acute sleep deprivation dysregulated Healthy controls (HC) were recruited, at the same time, they
the affective network (13–15), so it is not surprising that were matched with patients in gender, age, and education (see
neuromodulations are effective on insomnia. Although many Table 1). All participants declaimed to have taken any sleep-aid
clinical trials proved their safety, surgery is still impractical for drugs or psychotropic drugs. Both the CI group and the HC
patients suffering from diseases of mild symptoms, for example, group received the same clinical assessment, a session of taVNS
chronic insomnia (16). Transcutaneous auricular vagus nerve treatment, and fMRI scans at baseline. After that the CI group
stimulation (taVNS) belongs to the category of neuromodulation. received 4 weeks’ of taVNS treatment while the HC group did
A clinical trial has shown the efficacy of taVNS on CI (17), but the not receive any treatment.
underlying brain mechanism is still quite unclear.
Prefrontal cortex is more vulnerable to insomnia (18). Ethical Review and Registration
The dysfunction of PFC is one of the main pathological The study was reviewed by the Ethics Committee of Institute of
manifestations of insomnia (19–21), neuroimaging studies Acupuncture and Moxibustion under China Academy of Chinese
reveal that sleep deprivation severely damages the PFC and Medical Sciences (CACMS) and registered at the Chinese Clinical
reduces its ability of task execution and stimuli regulation Trial Registry (NO. ChiCTR-15007374).
(22). Reduced Amplitude of Low Frequency Fluctuation
(ALFF) was found widely in the frontal lobe in patients Transcutaneous Auricular Vagus Nerve
with insomnia, which indicated a lower neuroexcitability. Stimulation
Moreover, the aberrant ALFF is related to the duration and The electro-acupuncture stimulator (SDZ-IIB, Hwato brand,
severity of insomnia (19). Patients suffering from evening- made in Su zhou, China) was attached to the bilateral cymba
types insomnia even have a lower metabolism and a reduced conchae through electrodes on the skin surface (see Figure 1).
diurnal variation in PFC (23). Stimulating the peripheral Parameters were set according to previous studies of taVNS
branches of vagus would widely modulate the neuroexcitability (17, 27): Dilatational wave of 4/20 Hz and pulse width of 0.2 ms
through the projections from NTS to the forebrain and limbic ± 30%. Current intensity was adjusted according to each patient’s
system (24). subjective feeling. Each taVNS session lasted for 30 min, twice

Frontiers in Neurology | www.frontiersin.org 2 March 2022 | Volume 13 | Article 827749

He et al. taVNS Modulates the PFC in CI Patients

a day for 4 weeks, which is recommended by guidelines for Germany). The scanning parameters were as follows. In
short-term medications of insomnia (7, 8). functional images, the blood oxygen level-dependent gradient
Echo Planar Imaging (EPI) sequence was used. One hundred
Clinical Assessments and forty four volumes lasted 6 min 10 s, repeat time/echo time:
All participants accepted Pittsburgh Sleep Quality Index (PSQI) 2,500/30 ms, flip angle = 90 degrees, scanning field of view:
and Flinders Fatigue Scale (FFS) before and after the 4 weeks’ of 240 mm × 240 mm, matrix: 64 × 64, number of layers: 43,
treatment. To exclude the risk of depressive or anxiety symptoms, layer thickness/spacing: 3.0/1.0 mm. In high-definition structure
which may independently affect imaging findings, we used image, three-dimensional magnetization was used to prepare fast
Hamilton Rating Scale for Depression (HAMD) and Hamilton gradient echo sequence, repeat time/echo time: 2,500/2.98 ms,
Anxiety Rating Scale (HAMA) to estimate the mental status flip angle: 7 degree, field of view: 256 mm × 256 mm, matrix: 64
of all the participants. Before and after the taVNS treatment, × 64, number of layers: 48; Layer thickness/spacing: 1.0/1.0 mm.
the patient would be excluded with a total score of HAMD or
HAMA >7. The process of this study is shown in Figure 2.
FMRI Data Preprocessing
In addition, we screened all patients’ T2-weighted images and
DPABI (http://rfmri.org/DPABI) software (30), a SPM-based
structural images to ruled out most of the serious metabolic
functional MRI preprocessing pipeline, was used for data
or immune-related neuropsychiatric diseases, cerebrovascular
preprocessing. The preprocessing steps were as follows. Convert
diseases, inflammatory diseases of central never system, and
DICOM file into NIFTI. Remove the first 10 time points. The
intracranial tumors.
remaining 134 volumes were slice-time corrected and realigned
according to Friston 24-parameter model. The nuisance signals
MRI Data Acquisition (including linear trend, head-motion, signals of cerebrospinal
Rest-state functional magnetic resonance imaging (Rs-fMRI) fluid, and white matter) were regressed out from the data
were performed before and after the first 30 min’ taVNS (31). Then the functional images were co-registered to the
session. Participants were told to keep their eyes closed T1-weighted structural images, which were segmented through
and not fall asleep during the scan. The fMRI data was Voxel Based Morphometry (VBM). Derived images were
acquired by Siemens 3.0T Skyra equipment (Siemens; Munich, normalized to Montreal Neurological Institute (MNI) space
according to transformation parameters estimated by VBM.
The limitations of the signal-to-noise ratio and disputes in
sampling and preprocessing strategies for fMRI data, the existing
voxel based analysis studies are sometimes contradictory. To get a
better presentation of the short- time intervention, we employed
ALFF, fALFF, and ReHo to reveal the reproductive results.

FMRI Data Processing

The ALFF and fALFF were calculated from the normalized
images after smoothing (6 mm Gaussian kernel full width at half
maximum smooth nucleus) to the MNI space. ALFF value was
calculated as the average square root of the power spectrum range
of 0.01–0.1 Hz and converted to a frequency domain through the
FIGURE 1 | (A) The electrodes were attached to the surface of cymba fast Fourier transform process. FALFF value is the ratio of power
conchae. (B) The stimulating place of taVNS. RS-fMRI, rest-state functional in the specific frequency band of the whole detected frequency
Magnetic Resonance Imaging; taVNS, Transcutaneous Auricular Vagus
range. The ALFF and fALFF maps were also transferred to
Nerve Stimulation.
mean ALFF and fALFF maps by subtraction of the global mean

FIGURE 2 | Changes of fMRI induced by instant taVNS were measured before and after the first treatment on the two groups. Clinical scales for patients were
assessed before and after the 4 weeks’ taVNS treatment.

Frontiers in Neurology | www.frontiersin.org 3 March 2022 | Volume 13 | Article 827749

He et al. taVNS Modulates the PFC in CI Patients

value. The mean ALFF and fALFF values were converted to Z- TABLE 2 | Improvement of PSQI and FFS after 4-weeks taVNS treatment (x ± s).
distribution for standardization. Then we got the zALFF and
Items Baseline After treatment Z p-value
zfALFF map.
Regional homogeneity (ReHo) is calculated by voxel based on PSQI (N = 20) 12.7 ± 3.715 9.75 ± 4.278† 3.337 0.003
Kendall’s coefficient of concordance (KCC) for the time series of FFS (N = 20) 14.5 ± 5.92 11.5 ± 4.136* 2.860 0.010
a given voxel with its nearest neighbors (32). ReHo maps was
†
calculated through the unsmoothed and filtered (0.01–0.1 Hz) *p < 0.05; p < 0.01; Z, Wilcoxon rank testing; PSQI, Pittsburgh Sleep Quality Index; FFS,
images to remove physiological signals such as heartbeat and Flinders Fatigue Scale; Change at week 4 to baseline mean (95% CI).

respiration. Then ReHo maps were taken to mean ReHo maps

by subtraction of the mean voxel wise ReHo in the entire brain
and standardized into Z-value (zReHo Maps). Calculated zReHo First taVNS Session Activated the Similar
maps were smoothed to MNI space with 6 mm Gaussian kernel Location in Left dlPFC and Adjusted Its FC
full width at half maximum smooth nucleus at last. With PFC
FC is the Pearson’s correlations of the temporal fMRI signals The CI group showed lower ALFF and fALFF in dlPFC
between a Region of Interest (ROI) and all brain. Activated or and higher ReHo in Precuneus when compared to HC
deactivated regions found by the above voxel based analyses group (see Supplementary Material 1), which is similar to
would be used as the Region of Interest (ROI) for seed to voxel previous studies (19, 33). Three different voxel based analyses
FC analysis. FC were computed by voxel in the normalized showed consistent results. Namely, the first taVNS session
image after smoothing to Montreal Neurological Institute (MNI) up- regulated left dlPFC in the CI group (see Figures 3A,B
space (6 mm Gaussian kernel full width at half maximum smooth and Table 3). ALFF analysis showed the activation aroused by
nucleus). All images were band-pass filtered (0.01–0.1 Hz) taVNS was higher in CI group than in the HC group (see
before FC was computed. Pearson’s correlation coefficients were Figure 3C). Then the activated dlPFC found by the ALFF,
transformed into normally distributed scores according to the fALFF, and ReHo were merged as one ROI. The following
Fisher’s R- to -Z transformation. seed to voxel FC analysis revealed decreased FC between
dlPFC and bilateral dormedial prefrontal cortex (dmPFC) (see
Statistics Figures 3B,D).
In SPSS 25 (SPSS Inc., Chicago, IL, USA), two sample T-test
and χ2 tests were applied to compare the baseline characteristics A Lower ALFF or ReHo Value in dlPFC
between the CI and HC group. Paired T-test was applied to Before the First Session Correlating With
compare within group changes of PSQI and FFS scores in the Higher PSQI Score After 4 Weeks’ of
CI group. Treatment
For the fMRI images, the between group differences were When the correlations were examined between the clinical scales’
performed with independent two sample T-test, with an scores and the fMRI values, several significant results were
uncorrected p-value < 0.05. Paired T-tests were performed to defined. At baseline, ALFF values in dlPFC were negatively
determine the within-group differences in the group, before and correlated with the patients’ PSQI (R = −0.536, p < 0.01) and
after the first taVNS session. For the within group comparisons, FFS (R = −0.537, p < 0.05) score after 4 weeks’ of treatment. The
multiple comparison corrections were performance in Gaussian baseline ReHo values in dlPFC were also negatively correlated
random field correction (GRF), combined voxel wise p-value with the after-treatment PSQI (R = −0.545, p < 0.05) (see
< 0.001 with cluster p-value < 0.05 (two tailed). To clarify Figure 3E).
the behavioral associations of ALFF, fALFF, ReHo, and FC,
we performed Pearson correlation analyses between the fMRI
values and clinical scales in SPSS 25, controlling for age, sex, DISCUSSION
and education.
Our current study revealed that taVNS improved the CI
symptoms. In the first session, the taVNS up- regulated the left
dlPFC and reduced its FC with bilateral dmPFC. The baseline
RESULTS
ReHo and ALFF values in the left dlPFC were correlated with the
taVNS Improved PSQI and FFS Scores PSQI or FFS scores after 4 weeks’ of treatment.
Out of 24 patients, two were excluded, one because of stroke
history found by structural images and the other because dlPFC Is a Potential Targeting Brain Region
the patient was diagnosed with bipolar disorder. Another two of taVNS Treatment on CI
patients have withdrawn from the study. At last, twenty patients According to the hyperarousal theory, the Ascending Reticular
completed the 4 weeks’ of taVNS treatment as well as the two Activating System (ARAS) promotes the soberness of human
fMRI scan sessions. The mean duration of insomnia was 95.2 brain. Patients with CI have higher FC between the thalamus
months. Both PSQI (N = 20, p < 0.01 95%CI) and FFS (N = 20, and dlPFC, when compared to good sleepers. As a result, some
p < 0.05, 95%CI) improved after the 4 weeks’ taVNS treatment brain regions reduce their activity to compensate for the bottom-
(see Table 2). up arousal effects originating from ARAS. For example, dlPFC

Frontiers in Neurology | www.frontiersin.org 4 March 2022 | Volume 13 | Article 827749

He et al. taVNS Modulates the PFC in CI Patients

FIGURE 3 | (A) Different voxel based analyses showed similar up- regulated area in left dorsolateral prefrontal cortex. (B) Changes of different voxel based analyses
before and after taVNS. (C) ALFF analysis showed taVNS aroused higher activation in the CI group than in the HC group. (D) The FC between the up- regulated
dlPFC and bilateral dorsomedial prefrontal cortex decreased after taVNS. (E) Correlations between the clinical scales’ scores and the fMRI values. PSQI, Pittsburgh
Sleep Quality Index; FFS, Flinders Fatigue Scale; ALFF, amplitude of low frequency fluctuation; fALFF, Fractional ALFF; ReHo, Regional homogeneity; FC, Functional
connectivity; 0w, before taVNS treatment; 4w, after 4 week taVNS treatment; Relief Rate, The remission rate was defined as the difference in scale scores before and
after treatment divided by the scale scores before treatment. * p < 0.05; ** p < 0.01.

show a decreased ALFF in insomniacs (21), which is similar to answer to what we have observed. FC Maps between dlPFC
what we have found. DlPFC is a core region of cognitive control and subgenual cingulate has becoming a promising method for
network (CCN) (34–37), that is why insomniacs have lower navigating TMS in treating depression (39). Interestingly, our
working efficiency and they are vulnerable to fatigue despite of study also found that patients whose initial state of dlPFC was
their overexcited global status. low functioning would have a higher PSQI and FFS after 4 weeks
DlPFC is actually the most common stimulating target of of treatment. However, the limited sample size failed to reveal any
transcranial magnetic stimulation (TMS). TMS on dlPFC can correlation between the remission rate and ALFF values or ReHo
reduce the heart rate, the connection between the vagus and PFC values. A fMRI study reveals CBT increase the fALFF values
is the anatomical basis of these phenomena (38). Unlike TMS, in dlPFC and decrease the fALFF values in dmPFC in patient
taVNS activates dlPFC indirectly. The connection offers a potent with major depression disorder (40). This phenomenon indicates

Frontiers in Neurology | www.frontiersin.org 5 March 2022 | Volume 13 | Article 827749

He et al. taVNS Modulates the PFC in CI Patients

TABLE 3 | Brain changes after the first taVNS session (N = 20). Electroencephalogram (EEG) has a better time resolution than
fMRI. Many neuromodulations use EEG as a brain-machine
Items Brain regions BA Number MNI Peak
(AAL) of voxels coordinates(mm) intensity
interface to improve stimulating parameters. Our study found the
cortex is the most outstanding brain region affected by instant
X Y Z taVNS. It’s easier to get stable EEG signals of the cortex. Using
EEG to explore biomarkers of a certain neuromodulation would
ALFF Frontal_Mid_L 10/46 37 −39 36 21 7.791
Frontal_Inf_Tri_L
be of higher translational value than fMRI.
fALFF Frontal_Mid_L 10/46 26 −39 39 21 8.305
Frontal_Inf_Tri_L Limitations
ReHo Frontal_Mid_L 10/45/46 71 −39 39 21 7.364 First, there was no obvious decrease in PSQI in the current study,
Frontal_Inf_ which may be due to the fact that we only recruited patients with
Tri_L mild primary insomnia to ensure the consistency of the basic
FC Frontal_Sup_ 9/10 28 3 57 21 −5.143 state. Second, while our study revealed that the initial status of
Medial_R PFC in patients with insomnia was related to curative effect, no
Frontal_Sup_
Medial_L
difference were found between the changes of dlPFC and the
changes of the patients’ clinical scores. This may be due to the
ALFF, Low frequency fluctuation in the left dorsolateral prefrontal lobe; fALFF, Fractional limited sample size. Transient taVNS cannot completely explain
amplitude of low frequency fluctuation in the left dorsolateral prefrontal lobe; ReHo,
Regional homogeneity in the left dorsolateral prefrontal lobe; FC, Functional connectivity
the efficacy of mechanism of taVNS. The positive results we
between left dorsolateral prefrontal and medial prefrontal lobe; AAL, Anatomical Automatic found need a longer observation to get more convincing results.
Labeling; MNI, Montreal Neurological Institute; BA, Brodmann area. Third, a placebo control group is indeed the best designed control
groups of this study. At last, the sample size was too small and
we only studied the EPI sequence. Despite of the limitations, we
increased nervous excitability in the dlPFC, which is similar to
intend to provide the potential predicting imaging biomarkers for
what we have found after the first session of taVNS treatment. We
the suitable patients who are sensitive to taVNS.
speculate that taVNS and CBT may share a similar brain effect on
the dlPFC.
CONCLUSIONS
taVNS Lowered the CCN’s Monitoring to In this study, we found a short time taVNS aroused the left PFC in
Default Mode Network patients with insomnia. The changes of PFC could be replicated
Patients with CI have an abnormal FC between default through different voxel-based analyses. The projection from NTS
mode network (DMN) and the additional brain regions when to forebrain might be the anatomical basis of our findings.
compared with good sleepers (9, 41), which aggravate the
hyperarousal status of the brain. Increased ALFF values are found
in brain regions related to sensation and attention (19). That is DATA AVAILABILITY STATEMENT
why patients with insomnia are more sensitive to external stimuli
The raw data supporting the conclusions of this article will be
and are easier to be awakened. The FC within DMN, especially
made available by the authors, without undue reservation.
between the prefrontal lobe and the posterior DMN, decreases
when we fall asleep (19, 41). The dysfunctional DMN also leads
to the abnormal FCs within DMN, which impairs both the sleep ETHICS STATEMENT
structure and working memory (33, 42). DmPFC is one of the
most prominent brain regions of the abnormal frontal DMN The studies involving human participants were reviewed and
(41, 43). Study has also confirmed that dmPFC is the key area approved by Ethics Committee of Institute of Acupuncture and
for maintaining sleep (43). Patients with CI would pay excessive Moxibustion under China Academy of Chinese Medical Sciences.
attention to sleep quality, which would aggravates frustration The patients/participants provided their written informed
(44). This is because mPFC is connected with the hippocampus, consent to participate in this study. Written informed consent
amygdala, nucleus accumbens, and hypothalamus. They manage was obtained from the individual(s) for the publication of any
reward circuit and emotions (34, 45). potentially identifiable images or data included in this article.
Long-term sleep deprivation leads to a decompensated
Salient Network (SN). CCN should allocate more resources to AUTHOR CONTRIBUTIONS
compensate the loosed ability of SN to modulate the aberrant
DMN. The current study found that taVNS decreased the FC This article was written mainly by J-KH and B-HJ. The research
between left dlPFC and bilateral dmPFC, which is opposite to scheme was designed by P-JR. Patients were recruited and
the pathological changes of the patients with insomnia (2, 46), assessed by LL and J-LZ. FMRI data were collected by BZ and
indicating that CCN has lowered its monitoring to DMN, and were preprocessed by J-KH. Statistics and mapping were assisted
the excessive consumption of CCN reduced. We speculate that by J-KH, Z-GZ, and Y-ZB. S-YL, YW, J-LF, and P-JR reviewed
taVNS would alleviate the symptoms of low efficiency and fatigue the article. Text correction was done by M-ZW. All authors
in patients with CI. contributed to the article and approved the submitted version.

Frontiers in Neurology | www.frontiersin.org 6 March 2022 | Volume 13 | Article 827749

He et al. taVNS Modulates the PFC in CI Patients

FUNDING SUPPLEMENTARY MATERIAL

This study was supported by the National Key Research and The Supplementary Material for this article can be found
Development Program (No. 2018YFC1705800) and the National online at: https://www.frontiersin.org/articles/10.3389/fneur.
Natural Science Foundation (Nos. 81473780 and 81774433). 2022.827749/full#supplementary-material

REFERENCES 16. Johnson RL, Wilson CG. A review of vagus nerve stimulation as a therapeutic
intervention. J Inflamm Res. (2018) 11:203–13. doi: 10.2147/JIR.S163248
1. Buysse DJ. Insomnia. JAMA. (2013) 309:706–16. doi: 10.1001/jama.2013.193 17. Jiao Y, Guo X, Luo M, Li S, Liu A, Zhao Y, et al. Effect of transcutaneous
2. Lee YJG, Kim S, Kim N, Choi JW, Park J, Kim SJ, et al. Changes in subcortical vagus nerve stimulation at auricular concha for insomnia: a randomized
resting-state functional connectivity in patients with psychophysiological clinical trial. Evid Based Complement Altern Med. (2020) 2020:6049891.
insomnia after cognitive–behavioral therapy: changes in resting-state FC doi: 10.1155/2020/6049891
after CBT for insomnia patients. Neuroimage Clin. (2018) 17:115–23. 18. Verweij IM, Romeijn N, Smit DJA, Piantoni G, Van Someren EJW, van der
doi: 10.1016/j.nicl.2017.10.013 Werf YD. Sleep deprivation leads to a loss of functional connectivity in frontal
3. Licata SC, Lowen SB, Trksak GH, MacLean RR, Lukas SE. Zolpidem brain regions. BMC Neurosci. (2014) 15:88. doi: 10.1186/1471-2202-15-88
reduces the blood oxygen level-dependent signal during visual system 19. Li C, Ma X, Dong M, Yin Y, Hua K, Li M, et al. Abnormal spontaneous
stimulation. Prog Neuropsychopharmacol Biol Psychiatry. (2011) 35:1645–52. regional brain activity in primary insomnia: a resting-state functional
doi: 10.1016/j.pnpbp.2011.05.015 magnetic resonance imaging study. Neuropsychiatr Dis Treat. (2016) 12:1371–
4. Mi WF, Tabarak S, Wang L, Zhang SZ, Lin X, Du LT, et al. Corrigendum 8. doi: 10.2147/NDT.S109633
to: effects of agomelatine and mirtazapine on sleep disturbances in 20. Dai XJ, Nie X, Liu X, Pei L, Jiang J, Peng DC, et al. Gender differences in
major depressive disorder: evidence from polysomnographic and regional brain activity in patients with chronic primary insomnia: evidence
resting-state functional connectivity analyses. Sleep. (2021) 44:zsaa239. from a resting-state fMRI study. J Clin Sleep Med. (2016) 12:363–74.
doi: 10.1093/sleep/zsaa239 doi: 10.5664/jcsm.5586
5. Mysliwiec V, Martin JL, Ulmer CS, Chowdhuri S, Brock MS, Spevak C, 21. Zou Q, Ross TJ, Gu H, Geng X, Zuo XN, Hong LE, et al. Intrinsic resting-
et al. The management of chronic insomnia disorder and obstructive sleep state activity predicts working memory brain activation and behavioral
apnea: synopsis of the 2019 U.S. Department of Veterans Affairs and U.S. performance. Hum Brain Mapp. (2013) 34:3204–15. doi: 10.1002/hbm.
Department of Defense clinical practice guidelines. Ann Intern Med. (2020) 22136
172:325–36. doi: 10.7326/M19-3575 22. Chuah YML, Venkatraman V, Dinges DF, Chee MWL. The neural
6. Krystal AD, Prather AA, Ashbrook LH. The assessment and basis of interindividual variability in inhibitory efficiency after sleep
management of insomnia: an update. World Psychiatry. (2019) 18:337–52. deprivation. J Neurosci. (2006) 26:7156–62. doi: 10.1523/JNEUROSCI.0906-
doi: 10.1002/wps.20674 06.2006
7. Riemann D, Baglioni C, Bassetti C, Bjorvatn B, Dolenc Groselj L, Ellis JG, et al. 23. Hasler BP, Germain A, Nofzinger EA, Kupfer DJ, Krafty RT, Rothenberger
European guideline for the diagnosis and treatment of insomnia. J Sleep Res. SD, et al. Chronotype and diurnal patterns of positive affect and affective
(2017) 26:675–700. doi: 10.1111/jsr.12594 neural circuitry in primary insomnia. J Sleep Res. (2012) 21:515–26.
8. Qaseem A, Kansagara D, Forciea MA, Cooke M, Denberg TD, Barry MJ, doi: 10.1111/j.1365-2869.2012.01002.x
et al. Management of chronic insomnia disorder in adults: a clinical practice 24. Shouse MN, Langer J, Bier M, Farber PR, Alcalde O, Moghimi R, et al.
guideline from the American college of physicians. Ann Intern Med. (2016) The α2 adrenoreceptor agonist clonidine suppresses seizures, whereas
165:125–33. doi: 10.7326/M15-2175 the α2 adrenoreceptor antagonist idazoxan promotes seizures: pontine
9. Gong L, Yu S, Xu R, Liu D, Dai X, Wang Z, et al. The abnormal microinfusion studies of amygdala-kindled kittens. Brain Res. (1996) 731:203–
reward network associated with insomnia severity and depression in 7. doi: 10.1016/0006-8993(96)00594-X
chronic insomnia disorder. Brain Imaging Behav. (2021) 15:1033–42. 25. Fang J, Rong P, Hong Y, Fan Y, Liu J, Wang H, et al. Transcutaneous
doi: 10.1007/s11682-020-00310-w vagus nerve stimulation modulates default mode network in
10. Wang HX, Wang L, Zhang WR, Xue Q, Peng M, Sun ZC, et al. Effect major depressive disorder. Biol Psychiatry. (2016) 79:266–73.
of transcranial alternating current stimulation for the treatment of chronic doi: 10.1016/j.biopsych.2015.03.025
insomnia: a randomized, double-blind, parallel-group, placebo-controlled 26. Fang J, Egorova N, Rong P, Liu J, Hong Y, Fan Y, et al. Early
clinical trial. Psychother Psychosom. (2020) 89:38–47. doi: 10.1159/000504609 cortical biomarkers of longitudinal transcutaneous vagus nerve stimulation
11. Castillo PR, Middlebrooks EH, Grewal SS, Okromelidze L, Meschia JF, treatment success in depression. NeuroImage Clin. (2017) 14:105–11.
Quinones-Hinojosa A, et al. Globus pallidus externus deep brain stimulation doi: 10.1016/j.nicl.2016.12.016
treats insomnia in a patient with Parkinson disease. Mayo Clin Proc. (2020) 27. Liu J, Fang J, Wang Z, Rong P, Hong Y, Fan Y, et al. Transcutaneous
95:419–22. doi: 10.1016/j.mayocp.2019.11.020 vagus nerve stimulation modulates amygdala functional connectivity
12. Magdaleno-Madrigal VM, Valdés-Cruz A, Martínez-Vargas D, in patients with depression. J Affect Disord. (2016) 205:319–26.
Martínez A, Almazán S, Fernández-Mas R, et al. Effect of electrical doi: 10.1016/j.jad.2016.08.003
stimulation of the nucleus of the solitary tract on the development of 28. Tu Y, Fang J, Cao J, Wang Z, Park J, Jorgenson K, et al. A distinct
electrical amygdaloid kindling in the cat. Epilepsia. (2002) 43:964–9. biomarker of continuous transcutaneous vagus nerve stimulation
doi: 10.1046/j.1528-1157.2002.05702.x treatment in major depressive disorder. Brain Stimul. (2018) 11:501–8.
13. Baglioni C, Spiegelhalder K, Regen W, Feige B, Nissen C, Lombardo C, doi: 10.1016/j.brs.2018.01.006
et al. Insomnia disorder is associated with increased amygdala reactivity to 29. Riemann D. Hyperarousal and insomnia: state of the science. Sleep Med Rev.
insomnia-related stimuli. Sleep. (2014) 37:1907–17. doi: 10.5665/sleep.4240 (2010) 14:17. doi: 10.1016/j.smrv.2009.09.002
14. Wang T, Li S, Jiang G, Lin C, Li M, Ma X, et al. Regional homogeneity 30. Yan CG, Wang X Di, Zuo XN, Zang YF. DPABI: Data processing &
changes in patients with primary insomnia. Eur Radiol. (2016) 26:1292–300. analysis for (resting-state) brain imaging. Neuroinformatics. (2016) 14:339–51.
doi: 10.1007/s00330-015-3960-4 doi: 10.1007/s12021-016-9299-4
15. Huang Z, Liang P, Jia X, Zhan S, Li N, Ding Y, et al. Abnormal amygdala 31. Friston KJ, Williams S, Howard R, Frackowiak RSJ, Turner R. Movement-
connectivity in patients with primary insomnia: evidence from resting state related effects in fMRI time-series. Magn Reson Med. (1996) 35:346–55.
fMRI. Eur J Radiol. (2012) 81:1288–95. doi: 10.1016/j.ejrad.2011.03.029 doi: 10.1002/mrm.1910350312

Frontiers in Neurology | www.frontiersin.org 7 March 2022 | Volume 13 | Article 827749

He et al. taVNS Modulates the PFC in CI Patients

32. Zang Y, Jiang T, Lu Y, He Y, Tian L. Regional homogeneity 42. Drummond SPA, Walker M, Almklov E, Campos M, Anderson DE, Straus
approach to fMRI data analysis. Neuroimage. (2004) 22:394–400. LD. Neural correlates of working memory performance in primary insomnia.
doi: 10.1016/j.neuroimage.2003.12.030 Sleep. (2013) 36:1307–16. doi: 10.5665/sleep.2952
33. Marques DR, Gomes AA, Caetano G, Castelo-Branco M. Insomnia disorder 43. Koenigs M, Holliday J, Solomon J, Grafman J. Left dorsomedial frontal
and brain’s default-mode network. Curr Neurol Neurosci Rep. (2017) 18:45. brain damage is associated with insomnia. J Neurosci. (2010) 30:16041–3.
doi: 10.1007/s11910-018-0861-3 doi: 10.1523/JNEUROSCI.3745-10.2010
34. Kober H, Barrett LF, Joseph J, Bliss-Moreau E, Lindquist K, Wager TD. 44. Hamilton JP, Furman DJ, Chang C, Thomason ME, Dennis E, Gotlib IH.
Functional grouping and cortical-subcortical interactions in emotion: a Default-mode and task-positive network activity in major depressive disorder:
meta-analysis of neuroimaging studies. Neuroimage. (2008) 42:998–1031. implications for adaptive and maladaptive rumination. Biol Psychiatry. (2011)
doi: 10.1016/j.neuroimage.2008.03.059 70:327–33. doi: 10.1016/j.biopsych.2011.02.003
35. MacDonald A, Cohen J, Stenger VA, Carter C. Functional double dissociation 45. Casement MD, Keenan KE, Hipwell AE, Guyer AE, Forbes EE. Neural
of dorsolateral prefrontal cortex and anterior cingulate cortex in cognitive reward processing mediates the relationship between insomnia symptoms and
control. Neuroimage. (2000) 11:1835–8. doi: 10.1126/science.288.5472.1835 depression in adolescence. Sleep. (2016) 39:439–47. doi: 10.5665/sleep.5460
36. Niendam TA, Laird AR, Ray KL, Dean YM, Glahn DC, Carter CS. Meta- 46. Pang R, Zhan Y, Zhang Y, Guo R, Wang J, Guo X, et al. Aberrant functional
analytic evidence for a superordinate cognitive control network subserving connectivity architecture in participants with chronic insomnia disorder
diverse executive functions. Cogn Affect Behav Neurosci. (2012) 12:241–68. accompanying cognitive dysfunction: a whole-brain, data-driven analysis.
doi: 10.3758/s13415-011-0083-5 Front Neurosci. (2017) 11:259. doi: 10.3389/fnins.2017.00259
37. Cole MW, Schneider W. The cognitive control network: integrated
cortical regions with dissociable functions. Neuroimage. (2007) 37:343–60. Conflict of Interest: The authors declare that the research was conducted in the
doi: 10.1016/j.neuroimage.2007.03.071 absence of any commercial or financial relationships that could be construed as a
38. Iseger TA, van Bueren NER, Kenemans JL, Gevirtz R, Arns M. A potential conflict of interest.
frontal-vagal network theory for Major Depressive Disorder: implications
for optimizing neuromodulation techniques. Brain Stimul. (2020) 13:1–9. Publisher’s Note: All claims expressed in this article are solely those of the authors
doi: 10.1016/j.brs.2019.10.006 and do not necessarily represent those of their affiliated organizations, or those of
39. Cash RFH, Cocchi L, Lv J, Wu Y, Fitzgerald PB, Zalesky A. Personalized
the publisher, the editors and the reviewers. Any product that may be evaluated in
connectivity-guided DLPFC-TMS for depression: advancing computational
this article, or claim that may be made by its manufacturer, is not guaranteed or
feasibility, precision and reproducibility. Hum Brain Mapp. (2021) 42:4155–
72. doi: 10.1002/hbm.25330 endorsed by the publisher.
40. Shu Y, Kuang L, Huang Q, He L. Fractional amplitude of low-frequency
fluctuation (fALFF) alterations in young depressed patients with suicide Copyright © 2022 He, Jia, Wang, Li, Zhao, Zhou, Bi, Wu, Li, Zhang, Fang and Rong.
attempts after cognitive behavioral therapy and antidepressant medication This is an open-access article distributed under the terms of the Creative Commons
cotherapy: a resting-state fMRI study. J Affect Disord. (2020) 276:822–8. Attribution License (CC BY). The use, distribution or reproduction in other forums
doi: 10.1016/j.jad.2020.07.038 is permitted, provided the original author(s) and the copyright owner(s) are credited
41. Horovitz SG, Braun AR, Carr WS, Picchioni D, Balkin TJ, Fukunaga M, et al. and that the original publication in this journal is cited, in accordance with accepted
Decoupling of the brain’s default mode network during deep sleep. Proc Natl academic practice. No use, distribution or reproduction is permitted which does not
Acad Sci USA. (2009) 106:11376–81. doi: 10.1073/pnas.0901435106 comply with these terms.

Frontiers in Neurology | www.frontiersin.org 8 March 2022 | Volume 13 | Article 827749