The altered right frontoparietal network functional connectivity in migraine and the modulation effect of treatment

Abstract

Aims

This study aims to investigate the resting-state functional connectivity (rs-fc) of the right frontoparietal network (rFPN) between migraineurs and healthy controls (HCs) in order to determine how the rFPN rs-fc can be modulated by effective treatment.

Methods

One hundred patients and 46 matched HCs were recruited. Migraineurs were randomized to verum acupuncture, sham acupuncture, and waiting list groups. Resting-state functional magnetic resonance imaging data were collected before and after longitudinal treatments. Independent component analysis was applied in the data analysis.

Results

We found that migraineurs showed decreased rs-fc between the rFPN and bilateral precuneus compared with HCs. After treatments (real and sham), rFPN rs-fc with the precuneus was significantly reduced. This reduction was associated with headache intensity relief. In order to explore the role of the precuneus in acupuncture modulation, we performed a seed-based rs-fc analysis using the precuneus as a seed and found that the precuneus rs-fc with the bilateral rostral anterior cingulate cortex/medial prefrontal cortex, ventral striatum, and dorsolateral prefrontal cortex was significantly enhanced after treatment.

Conclusion

Our results suggest that migraineurs are associated with abnormal rFPN rs-fc. An effective treatment, such as acupuncture, may relieve symptoms by strengthening the cognitive adaptation/coping process. Elucidation of the adaptation/coping mechanisms may open up a new window for migraine management.

Keywords

Acupuncture migraine without aura fMRI independent component analysis frontoparietal network resting-state functional connectivity

Introduction

Migraine, a disabling chronic brain disorder, has become an important public healthcare and social issue due to its high prevalence, large medical burden (1), disabling effects (2), and serious reduction in quality of life (3). Nevertheless, many questions regarding its pathophysiology remain unclear. An improved understanding of the mechanisms underlying migraine and the modulating effects of treatment will open up new and promising avenues for the discovery of its causes, as well as the development of new therapeutic methods.

Despite its high prevalence and social burden, treatments for migraine are far from satisfactory (4). Previous studies showed that acupuncture may achieve its therapeutic effect by enhancing the human body’s self-regulation/healing process, and both verum or sham acupuncture can significantly relieve migraine headache intensity and frequency (5,6). Although the mechanism underlying acupuncture is complex, studies have demonstrated that non-specific effects (including attention, cognition, and expectation) play important roles (7). As a result, acupuncture treatment can be used as a useful treatment tool in order to investigate the underlying neural physiopathology of migraine.

Recently, investigators have begun to explore the functions of individual brain regions of interest as well as the neural networks involved in migraine (8 –12) and have found that the brain structure and function of migraineurs differ significantly from those of healthy individuals. These differences may result from their reaction to pain, light, and smell, or as an adaptive response to repeated pain (stressor) attacks (13,14).

More recently, accumulating evidence has suggested that resting-state functional connectivity (rs-fc) may be a valuable tool for understanding brain networks (15). One such brain network is the right frontoparietal network (rFPN), which plays an important role in cognitive control and top-down modulation (16), and has been reliably identified by independent component analysis (ICA) (17). For instance, previous studies suggested that pain and cognition interact reciprocally (i.e., pain can negatively influence cognitive performance, while cognition can significantly modulate our pain experience) (18,19). In addition, studies also show that attention can significantly influence pain experience through a top-down mechanism, and many studies demonstrated that distraction can reduce pain experience (20).

Recently, studies have shown that migraineurs are associated with impaired rFPN rs-fc (21,22). Therefore, this study aims to investigate: 1) the rFPN rs-fc between migraine patients (during the interictal period when they were free from headache symptoms) and matched healthy controls (HCs); and 2) how longitudinal acupuncture, which is an effective treatment, can modulate rFPN rs-fc in migraineurs. We hypothesize that migraineurs will be associated with altered rFPN rs-fc in order to adapt/cope with repeated headache attacks, while an effective non-pharmacological treatment can further modulate the adaptation/coping process.

Materials and methods

We briefly describe the experimental procedures below. Please also see previously published studies (23) for more details on the experimental procedure. The data have been used in previous studies investigating the functional connectivity of periaqueductal gray (PAG) changes between migraineurs and HCs, as well as the modulating effect of acupuncture. In this study, we used ICA in order to investigate the rFPN rs-fc difference between migraineurs and HCs and how an effective treatment can modulate the rFPN rs-fc. These results have not been reported before.

Participants

The Ethics Committee of the 1st Teaching Hospital of Chengdu University of Traditional Chinese Medicine approved all of the study procedures. The experiment was performed in accordance with approved guidelines. This study was then registered on ClinicalTrials.gov (NCT01152632, 27 June 2010). Patients (n = 100) and HCs (n = 46) were enrolled from the outpatient department of the 3rd Teaching Hospital, local advertisements, or the Chengdu University of Traditional Chinese Medicine campus. All participants signed a written consent form. Recruitment started in June 2011 and ended in November 2013.

Migraine without aura patients

Migraineur inclusion criteria were as follows: 1) 17–45 years of age (to improve demographic homogeneity) and right-handed; 2) matched the diagnosis of migraine without aura (MwoA) based on the International Classification of Headache Disorders, 2nd Edition (ICHD-II) MwoA criteria (24); 3) had not received any prophylactic headache medicine or acupuncture treatment in the past 3 months; 4) had experienced migraine for at least 6 months; and 5) had at least one headache attack per month in the past 3 months. Exclusion criteria included: 1) alcohol or drug abusers; 2) pregnant or lactating women; 3) suffered from psychiatric, neurologic, cardiovascular, respiratory, or renal illnesses; 4) had any other type of headache or a history of head trauma with loss of consciousness; 5) magnetic resonance imaging (MRI) contraindications such as claustrophobia; and 6) acupuncture contraindications such as excessive bleeding.

Healthy controls

Right-handed individuals aged between 17 and 45 years who were free from headache and other chronic pain conditions were recruited for this study as HCs. Each subject underwent a review of their medical history evaluation, physical examination, hepatic function, renal function, and routine analysis of blood, urine, and stool in order to exclude organic disease carriers. Individuals with abnormal test results or a history of head trauma with loss of consciousness, pregnancy, or lactation were excluded.

Study design

The total observation period for MwoA patients in this study was 8 weeks. After screening, all MwoA patients were randomized into five groups: verum acupuncture (VA) groups 1, 2, and 3 (VA1, VA2, and VA3), a sham acupuncture (SA) group, and the waiting list (WT) group. All patients were blinded to the group allocation. In this study, we included three VA prescriptions in order to better represent different acupoint selection strategies (25). Weeks 1–4 constituted the baseline phase, during which patients recorded baseline headache diaries. Weeks 5–8 constituted the intervention phase, during which patients in treatment groups received verum or sham acupuncture. All patients continued recording headache diaries during this treatment period. In addition, MRI scans were applied at the end of the fourth and eighth weeks for the migraine patients. All MwoA patients were migraine-free for at least 72 hours at the time of the MRI scan. HCs received only the baseline MRI scan (Figure 1).

Figure 1.

Study flow chart.

Interventions

Two licensed acupuncturists administered all acupuncture treatments. Acupoint and non-acupoint selection were similar to those in our previous randomized controlled trials (6). Acupoints selected in VA1 included Yanglingquan (GB34), Qiuxu (GB40), and Waiguan (SJ5). VA2 acupoints included Xiyangguan (GB33), Diwuhui (GB42), and Sanyangluo (SJ8). VA3 acupoints included Zusanli (ST36), Chongyang (ST42), and Pianli (L16). SA acupoints included non-acupoints 1, 2, and 3 (NAP1, NAP2, and NAP3) (Figure 2).

Figure 2.

Acupoint locations.

All acupoints and non-acupoints were punctured bilaterally using disposable needles. The needles were inserted perpendicularly at a penetration of 5–15 mm and were gently twisted, lifted, and thrust at an even amplitude, force, and speed to acquire deqi sensation (deqi sensation is a complex feeling including soreness, numbness, heaviness, distention, or dull pain at the site of needle placement) (26) in all treatment groups. The MwoA patients in acupuncture groups (AGs) received 20 treatments (30 minutes each) over a 4-week period: once per day for 5 weekdays followed by a 2-day break. Acupuncture or SA treatment was not performed on HCs and migraineurs in the WT group.

MwoA patients were instructed and agreed not to take any regular medications for migraine treatment. In cases of severe pain, ibuprofen (300 mg per capsule with sustained release) was allowed as a rescue medication.

Outcome measures

The clinical outcomes included headache intensity (visual analog scale (VAS) of 0–10) and frequency (number of migraines separated by pain-free intervals of at least 48 hours) in the past month obtained from the headache diary. The metrics obtained were consistent with the guidelines of the International Headache Society (IHS) for Clinical Trials in Migraine (27). In addition, the self-rating anxiety scale (SAS) and self-rating depression scale (SDS) were applied in order to assess the MwoA patients’ emotional status (28,29).

MRI data acquisition

MRI data were acquired with a 3.0 T magnetic resonance scanner (Siemens 3.0 T Trio Tim, Munich, Germany) with an eight-channel head coil at the West China Hospital MRI center. Prior to the functional run, a high-resolution structural image for each subject was acquired using a three-dimensional MRI sequence with a voxel size of 1 mm³ employing an axial fast spoiled gradient recalled sequence (repetition time (TR) = 1900 ms; echo time (TE) = 2.26 ms; data matrix: 256 × 256; field of view: 256 × 256 mm²). The blood oxygen level dependent (BOLD) resting-state functional images were obtained with echo-planar imaging (30 continuous slices with a slice thickness of 5 mm; TR = 2000 ms; TE = 30 ms; flip angle: 90°; total volumes: 180; in-plane resolution: 3.75 × 3.75 mm²; matrix size: 64 × 64; field of view: 256 × 256 mm²). Subjects were instructed to stay awake and remain motionless during the scan with their eyes closed.

Data analysis

Clinical data analysis

The clinical variables were analyzed using SPSS16.0 software (SPSS Inc., Chicago, IL). A threshold of p < 0.05 (two-tailed) was applied. Continuous variables were presented as the mean (standard deviation) with 95% confidence intervals. Categorical variables were described as n (percentage). χ² was applied for categorical variable comparisons. For continuous variables, paired t-tests were applied for within-group comparisons, two-sample t-tests were applied for two-group comparisons, and one-way analysis of variance was applied when there were more than three groups.

ICA for resting-state functional MRI data

Resting state data were processed using FMRIB Software Library (FSL) and Analysis of Functional NeuroImages, software (AFNI) following the same processing steps (pipeline) described previously (17). Previous studies suggested ICA to be a reliable method for analyzing rs-fc, particularly for the attention and control networks (30,31).

Preprocessing of functional images included removal of non-brain structures, motion correction, temporal bandpass filtering at 0.01–0.1 Hz, spatial smoothing (6 mm full-width at half-maximum Gaussian kernel), and eight-parameter nuisance signal (global signal not included (32)) extraction. Similarly to a previous study (17), in order to co-register functional MRI (fMRI) images to a standard space, functional images were first registered to each individual’s high-resolution T1 anatomical scan and further registered to the Montreal Neurological Institute 152 (MNI152) template using affine transformations with 12 degrees of freedom (33,34).

Probabilistic ICA at low dimensionality (20 components) was performed (MELODIC and FSL (35)) in order to derive the group’s (n = 190) resting-state networks. Spatial correlations between our group-level networks and the template networks derived from 1414 healthy subjects (17) were calculated. The group-derived network that showed the highest spatial overlap with the rFPN in the template network was assigned to the rFPN.

Then, a dual-regression (36) analysis was applied. Using the rFPN as spatial regressors in a general linear model (GLM), we were able to extract the temporal dynamics associated with each spatial map. The resulting time courses served as temporal regressors in a GLM in order to generate subject-specific maps of the whole brain for each subject. Finally, group analyses were performed using whole-brain subject-specific network maps from the second GLM. The results represent the strength of the rs-fc for each voxel with the rFPN. A two-sample t-test and paired t-test were conducted in order to compare the between-group and within-group differences.

In addition, we also applied multiple regression analyses on migraine patients in order to explore the association between baseline functional connectivity and the corresponding migraine intensity as measured by VAS, as well as the association between pre- and post-treatment functional connectivity changes and corresponding VAS changes in all treatment groups, including age, gender, disease duration, SAS, and SDS as non-interest covariates. A threshold of z > 2.3 and p < 0.05 corrected was applied.

Seed-based functional connectivity

Using ICA analysis, we found that the precuneus plays an important role in the modulation of acupuncture. In order to further explore the modulation process, we performed a seed-based rs-fc analysis using the precuneus as a seed. The fMRI data were preprocessed using Data Processing Assistant for Resting-State fMRI (DPARSF) software (available at: http://rfmri.org/DPARSF). The fMRI images were slice timing and head motion corrected, co-registered to the respective structural images for each subject, segmented, regressed out eight-parameter nuisance signal (including six rigid body motion, white matter and cerebrospinal fluid signal), normalized using structural image-unified segmentation, and then re-sampled to 3-mm cubic voxels. After linear detrending, data were filtered using a typical temporal bandpass (0.01–0.08 Hz) in order to remove low-frequency noise (including slow scanner drifts) and the influences of higher frequencies reflecting cardiac and respiratory signals. We removed frames with framewise displacement (FD) > 0.5 mm (‘scrubbing’), and one time point before ‘bad’ time points and two time points after ‘bad’ time points were deleted. Finally, the data were smoothed using a full-width half maximum of 6 mm.

Functional connectivity analysis for individual subjects was carried out in DPARSF by applying a seed-region approach using the right precuneus (2, −62, 50, 3 mm) so that the coordinate represents the peak of the overlap cluster observed in the ICA results. Next, the averaged time course was obtained from the seed and correlation analysis was performed in a voxel-wise manner in order to generate the functional connectivity (FC) map. The correlation coefficient map was converted into a Fisher z-map using Fisher’s r-to-z transform by calling functions in REST (Resting-State fMRI Data Analysis Toolkit, software) to improve normality. Group analysis was calculated with a random effect model using SPM8 (Statistical Parametric Mapping 8, software). We first compared the rs-fc difference between MwoA patients and HCs using two-sample t-tests. Then, we compared the changes in rs-fc differences (post-treatment minus pre-treatment) between AGs (verum + sham) and the WT group in the factorial design module in SPM8. Similarly to previous study (23), a threshold of voxel-wise p < 0.005 (uncorrected) and p < 0.05 family-wise error corrected at cluster level was applied for all of the analyses.

Results

One hundred and fifty patients were screened, of which 100 patients were recruited for this study. Forty-six age- and gender-matched HCs were also recruited. Eighty-eight patients participated in the first fMRI scan and 81 patients participated in the second. Seven patients did not participate in the second fMRI scan due to scheduling conflicts (two in VA1, two in VA2, one in VA3, and two in the SA group). Of the 81 patients who participated in the two MRI scans, nine patients were excluded from the data analysis due to incomplete scans (lack of resting-state MRI or T1 anatomy; three in V1, one in V2, two in V3, two in the SA group, and one in the WT group) (Figure 1).

Baseline characteristics

We found no statistical differences between the VA1, VA2, VA3, SA, and WT groups in terms of age, sex, weight, height, duration of disease, headache intensity (VAS score), headache frequency, SAS, and SDS (p > 0.05). There were no statistical differences in age, gender, weight, and height between MwoA patients and HCs (Table 1).

Table 1.

Baseline characteristics of migraine without aura patients (subjects finished the two scans with completed data) in different groups and healthy controls.

Characteristics	VA1, n = 12	VA2, n = 14	VA3, n = 15	SA, n = 13	WT, n = 18	p-value*	MwoA, n = 72	HCs, n = 46	p-value**
Female, n (%)	10 (83.3%)	10 (71.4%)	12 (80.0%)	11 (84.6%)	14 (77.8%)	0.925	57 (79.2%)	36 (78.3%)	0.907
Age (years), mean (95% CI)	21.75 (20.70–22.80)	20.93 (19.74–22.12)	20.87 (19.89–21.85)	21.38 (20.75–22.02)	21.61 (20.54–22.68)	0.605	21.30 (20.89–21.73)	21.24 (20.98–21.50)	0.789
Height (cm), mean (95% CI)	159.0 (154.50–163.40)	161.93 (157.52–166.24)	159.1 (155.13–163.01)	157.9 (154.84–161.00)	162.3 (157.91–166.76)	0.390	160.22 (158.49–161.96)	161.11 (158.49–161.96)	0.493
Weight (kg), mean (95% CI)	52.50 (48.09–56.91)	55.21 (49.23–61.20)	50.17 (47.64–52.70)	50.23 (45.96–54.50)	53.95 (49.01–58.88)	0.375	52.49 (50.56–54.42)	51.13 (49.33–52.93)	0.335
Duration (months), mean (95% CI)	61.67 (44.38–78.95)	69.57 (56.10–100.19)	62.27 (42.79–82.54)	58.69 (34.29–64.63)	64.73 (58.52–95.81)	0.140	–	–	–
Headache intensity, mean (95% CI)	5.33 (4.46–6.20)	5.32 (4.77–5.87)	5.80 (5.20–6.40)	5.46 (4.52–6.40)	5.58 (5.14–6.03)	0.799	–	–	–
Headache frequency, mean (95% CI)	5.75 (3.89–7.61)	7.50 (5.61–9.39)	5.73 (3.58–7.88)	6.00 (4.04–7.96)	4.50 (3.07–5.93)	0.164	–	–	–
SAS score, mean (95% CI)	44.63 (39.74–49.51)	44.86 (40.36–49.35)	47.02 (41.69–52.34)	47.21 (42.23–52.20)	47.10 (41.92–52.26)	0.885	–	–	–
SDS score, mean (95% CI)	42.13 (35.01–49.24)	48.98 (43.97–53.99)	44.80 (38.22–51.38)	45.06 (39.25–50.86)	46.88 (42.01–51.74)	0.516	–	–	–

χ² test was applied for gender comparison, one-way analysis of variance was applied for the rest of the comparisons among VA1, VA2, VA3, SA, and WT groups; **χ² test was applied for gender comparison, two-sample t-tests were applied for the rest of the comparisons between MwoA patients and HCs.

HC: healthy control; MwoA: migraine without aura; VA: verum acupuncture; SA: sham acupuncture; SAS: self-rating anxiety scale; SDS: self-rating depression scale; WT: waiting list; CI: confidence interval.

Clinical outcomes

Compared with the baseline condition, all three VA groups (VA1, VA2, and VA3) showed improvements in headache intensity and headache frequency (p < 0.05) after treatment. Participants in the SA group only showed improvements in SAS and SDS (p < 0.05) and a trend for reduced headache intensity (p = 0.083) (Table 2). As expected, we found no significant differences between the VA1, VA2, VA3, and SA groups in headache intensity, headache frequency, SAS, and SDS improvements (p > 0.05), which is consistent with previously published meta-analysis reports (37,38), indicating that the specific effect of acupuncture treatment as compared with sham is only moderate. Because both verum and SA treatments reduced the headache intensity in migraine patients, we merged all AGs and found that AGs showed significantly greater improvements than the WT group in headache intensity and frequency (p < 0.05) (Table 2). Since the aim of this study is to use acupuncture treatment as a mediator in order to investigate the neural mechanism of migraine development, we merged all acupuncture treatment groups (verum and sham) in the following rs-fc analyses in order to investigate the modulating effect of an effective treatment.

Table 2.

Clinical outcomes before and after treatment in different groups.

Outcome measures	VA1, n = 12	VA2, n = 14	VA3, n = 15	SA, n = 13	WT, n = 18	AG, n = 54
Headache intensity, mean (95% CI)
Baseline	5.33 (4.46–6.20)	5.32 (4.77–5.87)	5.80 (5.20–6.40)	5.46 (4.52–6.40)	5.58 (5.14–6.03)	5.49 (5.15–5.83)
End of treatment	3.25 (2.39–4.11)	3.57 (2.83–4.31)	2.87 (2.01–3.72)	4.19 (3.34–5.04)	5.53 (4.69–6.36)	3.45 (3.06–3.85)
p-value	0.003	0.000	0.000	0.083	0.869	0.000
Headache frequency, mean (95% CI)
Baseline	5.75 (3.89–7.61)	7.50 (5.61–9.39)	5.73 (3.58–7.88)	6.00 (4.04–7.96)	4.50 (3.07–5.93)	6.26 (5.34–7.18)
End of treatment	3.92 (2.37–5.46)	5.93 (4.38–7.47)	4.20 (2.42–5.98)	6.15 (4.03–8.27)	8.17 (5.81–10.52)	5.06 (4.21–5.90)
p-value	0.014	0.021	0.013	0.895	0.001	0.003
SAS score, mean (95% CI)
Baseline	44.63 (39.74–49.51)	44.86 (40.36–49.35)	47.02 (41.69–52.34)	47.21 (42.23–52.20)	47.10 (41.92–52.26)	46.00 (43.71–48.23)
End of treatment	38.13 (31.15–45.10)	41.45 (37.12–45.78)	37.98 (32.61–43.36)	37.37 (33.09–41.64)	41.39 (37.83–44.94)	38.76 (36.36–41.17)
p-value	0.055	0.117	0.005	0.019	0.017	0.000
SDS score, mean (95% CI)
Baseline	42.13 (35.01–49.24)	48.98 (43.97–53.99)	44.80 (38.22–51.38)	45.06 (39.25–50.86)	46.88 (42.01–51.74)	45.35 (42.51–48.20)
End of treatment	40.63 (32.17–49.08)	42.68 (37.07–48.28)	39.15 (33.41–44.89)	37.56 (32.06–43.06)	41.18 (36.57–45.80)	40.01 (37.14–42.88)
p-value	0.518	0.040	0.023	0.049	0.066	0.000

Paired t-tests were applied for comparisons in each group.

AG: acupuncture group, including both verum acupuncture and sham acupuncture; VA: verum acupuncture; SA: sham acupuncture group; SAS: self-rating anxiety scale; SDS: self-rating depression scale; WT: waiting list; CI: confidence interval.

ICA results

As expected, ICA analysis including all subjects/scans has produced a rFPN, which mainly covers the right dorsolateral prefrontal cortex (DLPFC), ventrolateral prefrontal cortex (VLPFC), bilateral inferior/superior parietal lobules, medial prefrontal cortex (MPFC), insula, and left cerebellum (Figure 3 and Table 3). This component is consistent with findings from previous studies (39).

Figure 3.

Right frontoparietal network revealed by independent component analysis in this study.

Table 3.

Comparisons of the therapeutic effects between different groups.

Outcome measures	VA1, n = 12	VA2, n = 14	VA3, n = 15	SA, n = 13	p-value*	WT, n = 18	AG, n = 54	p-value**
Headache intensity, mean (95%CI)
End of treatment	3.25 (2.39–4.11)	3.57 (2.83–4.31)	2.87 (2.01–3.72)	4.19 (3.34–5.04)	0.101	5.53 (4.69–6.36)	3.45 (3.06–3.85)	0.000
End minus baseline	−2.08 (−3.31 to −0.86)	−1.75 (−2.42 to −1.08)	−2.93 (−4.01 to −1.86)	−1.27 (−2.73 to −0.19)	0.139	−0.06 (−0.75–0.64)	−2.04 (−2.57 to −1.50)	0.000
Headache frequency, mean (95%CI)
End of treatment	3.92 (2.37–5.46)	5.93 (4.38–7.47)	4.20 (2.42–5.98)	6.15 (4.03–8.27)	0.132	8.17 (5.81–10.52)	5.06 (4.21–5.90)	0.016
End minus baseline	−1.83 (−3.21 to −0.46)	−1.57 (−2.87 to −0.28)	−1.53 (−2.69 to −0.37)	0.15 (−2.32–2.63)	0.250	3.67 (1.63–5.70)	−1.20 (−1.96 to −0.44)	0.000
SAS score, mean (95%CI)
End of treatment	38.13 (31.15–45.10)	41.45 (37.12–45.78)	37.98 (32.61–43.36)	37.37 (33.09–41.64)	0.626	41.39 (37.83–44.94)	38.76 (36.36–41.17)	0.257
End minus baseline	−6.50 (−13.16–0.17)	−3.41 (−7.79–0.97)	−9.03 (−14.78 to −3.28)	−9.85 (−17.78 to −1.92)	0.379	−5.71 (−10.26 to −1.16)	−7.21 (−10.08 to −4.33)	0.591
SDS score, mean (95%CI)
End of treatment	40.63 (32.17–49.08)	42.68 (37.07–48.28)	39.15 (33.41–44.89)	37.56 (32.06–43.06)	0.638	41.18 (36.57–45.80)	40.01 (37.14–42.88)	0.675
End minus baseline	−1.50 (−6.44–3.44)	−6.30 (−12.26 to −0.35)	−5.65 (−10.38 to −0.92)	−7.50 (−14.96 to −0.05)	0.469	−5.69 (−11.81–0.43)	−5.34 (−5.34 to −8.03)	0.902

One-way analysis of variance was applied for the comparisons between the VA1, VA2, VA3, SA and WT groups; **two-sample t-tests were applied for the comparisons between MwoA patients and HCs.

AG: acupuncture group; HC: healthy control; VA: verum acupuncture; SA: sham acupuncture; SAS: self-rating anxiety scale; SDS: self-rating depression scale; WT: waiting list; CI; confidence interval.

Compared with HCs, MwoA patients showed significantly decreased rFPN rs-fc with the bilateral precuneus, lingual gyrus, middle temporal gyrus, and superior temporal gyrus; left fusiform and secondary somatosensory cortex (S2); and right cerebellum, inferior occipital gyrus, inferior temporal gyrus and cuneus (Figure 4 and Table 4). Regression analysis showed that the rFPN rs-fc with the bilateral precuneus, rostral anterior cingulate cortex (rACC)/MPFC, and left superior frontal gyrus was positively associated with migraine headache intensity (VAS) at the baseline. The rFPN rs-fc of the right inferior temporal gyrus was negatively associated with headache intensity (Figure 4 and Table 4).

Figure 4.

rFPN resting-state functional connectivity (rs-fc) results. (a) MwoA showed reduced rFPN rs-fc with the precuneus at baseline compared with HCs. (b) rFPN rs-fc between the precuneus and medial prefrontal cortex/rostral anterior cingulate cortex was associated with headache VAS at baseline. (c) Reduced rFPN rs-fc with the precuneus due to acupuncture treatment (verum + sham). (d) The greater reduction of rFPN rs-fc with the right precuneus was associated with greater headache intensity relief. A threshold of z > 2.3 and p < 0.05 was applied.

Table 4.

The right frontoparietal network resting-state functional connectivity (independent component analysis).

rFPN resting-state functional connectivity differences between MwoA patients and HCs at baseline							Association between VAS and rFPN resting-state functional connectivity in MwoA patients at baseline
Contrast	Voxels	Brain region	MNI (x, y, z)			z	Contrast	Voxels	Brain region	MNI (x, y, z)			z
HC > MwoA	884	L/R precuneus	−6	−58	46	4.12	Positive	358	L/R precuneus	4	−60	54	3.78
	595	L S2	−54	−32	22	4.17		349	L sup frontal G	−14	68	10	4.14
	689	L lingual G	−10	−98	−16	4.54		191	R/L pgACC/MPFC	−4	44	0	3.54
	829	L fusiform G Mid/sup temporal G	−46	−52	−10	4.48
	426	R lingual/cuneus	18	−84	4	4.17
	1086	R sup/mid/inf temporal G	38	−74	−4	4.57
	273	R cerebellum	40	−40	−28	4.72
MwoA > HC	No brain region above the threshold						Negative	162	R inf temporal G	52	−14	−32	4.22
Changes of rFPN resting-state functional connectivity in MwoA patients before and after acupuncture treatment							Association between VAS changes (post minus pre) and corresponding rFPN functional connectivity changes in MwoA patients
Contrast	Voxels	Brain region	MNI (x, y, z)			z	Contrast	Voxels	Brain region	MNI (x, y, z)			z
Pre > post	569	R precuneus	22	−62	40	4.02	Positive	212	L/R precuneus	−2	−70	34	3.39
Pre > post	195	L mid frontal G	−38	34	14	3.59	Positive	268	R postcentral G	14	−34	72	3.87
Post > pre	183	R/L dorsal PCC	10	−38	24	3.74	Negative	No brain region above the threshold

AG: acupuncture group; G: gyrus; HC: healthy control; Inf: inferior; L: left side; MwoA: migraine without aura; Mid: middle; PCC: posterior cingulate cortex; R: right side; rFPN: right frontoparietal network; MIN: Montreal Neurological Institute coordinate space; MPFC: medial prefrontal cortex; pgACC: perigenual cingulate cortex; Sup: superior; S2: secondary somatosensory cortex; VAS: visual analog scale.

Paired t-tests showed that after longitudinal treatment (real and sham), the rFPN rs-fc with the right precuneus and left middle frontal gyrus was significantly reduced; the rFPN rs-fc with the bilateral posterior cingulate cortex was significantly increased (Figure 4 and Table 4). Regression analysis showed that after treatment, the decrease of rFPN rs-fc with the bilateral precuneus, right paracentral gyrus, and postcentral gyrus was positively associated with a decrease in headache intensity (Figure 4 and Table 4).

In the WT group, we found that the rFPN rs-fc with the left inferior frontal gyrus (peak MNI coordinates: −52, 30, 8; cluster size: 351) and bilateral MPFC (peak MNI coordinates: 0, 44, 28; cluster size: 273) was significantly reduced; the rFPN rs-fc with the middle temporal gyrus/angular gyrus (peak MNI coordinates: −52, −68, 38; cluster size: 208) was significantly increased in MwoA patients (second time minus first time). The comparison between the acupuncture treatment (real and sham; post- minus pre-treatment) and the WT group showed a significant difference in the left cerebellum (peak MNI coordinates: −6, −46, −30; cluster size: 163) and right middle frontal gyrus (peak MNI coordinates: 38, 6, 46; cluster size: 123). In addition, we also compared the rFPN rs-fc change (post- minus pre-treatment) between SA and WT groups. The results showed a significant difference at the right precuneus (peak MNI coordinates: 24, −66, 26; cluster size: 205), right posterior cingulate cortex (peak MNI coordinates: 6, −60, 18; cluster size: 87), right cuneus (peak MNI coordinates: 8, −80, 40; cluster size: 35), and left cuneus (peak MNI coordinates: −4, −76, 26; cluster size: 20).

Seed-based rs-fc analysis results

We found that the right precuneus plays an important role in the pathophysiology of migraine, specifically finding: 1) reduced rs-fc between the rFPN and the right precuneus in MwoA patients as compared with HCs; 2) the rs-fc between the rFPN and right precuneus was associated with headache intensity in baseline MwoA patients; and 3) the decrease of rFPN rs-fc with the right precuneus was positively associated with a decrease in headache intensity due to acupuncture treatment. In order to further explore the role of the right precuneus in the pathology of migraine and its role in treatment, we applied a functional connectivity analysis using the peak of the overlapped cluster at the right precuneus (2, −62, 50, 3 mm) as the seed. We found that, compared with HCs, MwoA patients showed reduced right precuneus rs-fc with the left precuneus, supramarginal gyrus, and inferior temporal gyrus. After longitudinal acupuncture treatment (verum + sham), MwoA patients showed increased right precuneus rs-fc with the bilateral rACC/MPFC, ventral striatum, middle/inferior occipital gyrus, cuneus, DLPFC and cerebellum, and left VLPFC and right superior temporal gyrus (Figure 5 and Table 4).

Figure 5.

Precuneus resting-state functional connectivity (rs-fc) results. (a) MwoA showed reduced precuneus rs-fc with the supramarginal gyrus and inferior temporal gyrus compared with HCs. (b) MwoA patients showed increased right precuneus rs-fc with the bilateral rACC/MPFC, ventral striatum, DLPFC, and VLPFC due to acupuncture treatment (verum + sham). A threshold of voxel-wise p < 0.005 and p < 0.05 family-wise error corrected was applied.

Discussion

In this study, we found that MwoA patients showed reduced rFPN rs-fc with the bilateral precuneus during the interictal period compared with HCs. The rFPN rs-fc between the MPFC/rACC and precuneus was positively associated with headache intensity in MwoA patients at baseline. Interestingly, acupuncture (both real and sham), could significantly reduce rFPN rs-fc with the precuneus. Greater headache intensity relief was associated with greater reduced rFPN rs-fc with the right precuneus. Seed-based functional connectivity analysis (using the right precuneus as the seed) showed that longitudinal acupuncture treatment significantly enhanced the precuneus rs-fc with the MPFC/rACC, ventral striatum (40), and DLPFC/VLPFC (41,42).

A previous study suggested that the left frontoparietal network corresponds strongly to cognition–language paradigms, while the rFPN corresponds strongly to perception–somesthesis–pain (39). Thus, in this study, we only focused on the rFPN. Our results are partly consistent with previous studies in which investigators found impaired rFPN rs-fc in MwoA patients (21,22). In a previous study, Xue and colleagues (21) found that migraineurs (n = 23) showed increased rFPN rs-fc in the middle frontal gyrus and anterior insula. In another study, Russo and colleagues (22) found that MwoA patients (n = 14) had reduced rFPN rs-fc with the middle frontal gyrus and dorsal anterior cingulate cortex compared to HCs. The differences in rFPN rs-fc changes in migraineurs between our study and previous studies may not necessarily be contradictory, as we believe that these differences reflect the complexity of the neural physiopathology of migraine (43).

Studies have suggested that key regions of the rFPN, including the VLPFC, DLPFC, and the parietal gyrus (39), are involved in the cognitive control of pain, and that all of these regions have direct connections with the brain regions involved in affective (i.e., anterior cingulate cortex, MPFC, and amygdala) and sensory (i.e., primary somatosensory cortex (SI) and S2/insula) pain process components (41,42). We found that MwoA patients showed reduced rFPN rs-fc with the precuneus, which is associated with reduced headache intensity. In a previous study, investigators divided the precuneus into three sub-regions based on the rs-fc patterns: the anterior, central, and posterior precuneus. The precuneus region in our study is predominantly located in the central precuneus, which is connected to the dorsolateral/dorsomedial prefrontal and multimodal lateral inferior parietal cortices, suggesting a cognitive/associative brain function (44). Our results are consistent with previous studies showing that the precuneus is closely connected with the rFPN both anatomically and functionally (44,45).

The precuneus is a key node in the default mode network (DMN) (39), a brain network that is associated with self-referential processing (46,47) and mind-wandering (48), whereas the rFPN plays an important role in attention, memory processes, and cognitive control (16). We thus speculate that the reduced rs-fc between the rFPN and the precuneus in migraine patients before treatment may represent the brain’s self-compensatory adaptation/coping responses to continued attacks of migraine (20,49,50). Specifically, paying attention to pain (51) results in more suffering, while distraction (52) from pain results in a less painful experience. The body may automatically apply a distraction strategy as a natural response in order to avoid suffering.

Interestingly, we found that longitudinal acupuncture treatment could further reduce rFPN rs-fc with the right precuneus, and that reduced right precuneus–rFPN rs-fc was associated with migraine headache intensity reduction. Previous studies suggested that expectation, attention, and reappraisal are crucial components of the non-specific effects of treatments (41,42,53). In addition, studies have also shown that the non-specific effects play important roles in acupuncture’s treatment of chronic pain (54 –56). Taken together, our results suggest that acupuncture may relieve headache intensity by enhancing the self-compensatory adaptation/coping process. This finding is consistent with a previous brain imaging study of cognitive–behavioral therapy in fibromyalgia showing that, rather than reducing pain responses in patients with fibromyalgia, cognitive–behavioral therapy increases access to executive regions for the reappraisal of pain (57).

Seed-based functional connectivity analysis showed that after longitudinal acupuncture treatments, the right precuneus rs-fc with the MPFC/rACC was significantly increased. The MPFC/rACC has wide functional connections (58) and involvement in many functions; several studies have suggested that the MPFC/rACC is functionally connected with the PAG (Table 5) (59,60), a key region in the descending pain modulatory regions. Previous studies showed that acupuncture can increase rs-fc between the PAG and MPFC in patients with migraine (23) and knee osteoarthritis (61). We thus speculate that the increased precuneus rs-fc with the MPFC/rACC implies an enhancement between the self-referral system and the pain modulation process.

Table 5.

Results from the seed-based functional connectivity analysis, showing the altered right precuneus resting-state functional connectivity in MwoA as compared with healthy controls and the pre- and post-treatment resting-state functional connectivity differences across patients.

R precuneus resting-state functional connectivity differences between MwoA patients and HCs
Contrast	Voxels	Brain region	MNI (x, y, z)			z
HC > MwoA	322	L precuneus	42	−63	39	4.68
		L supramarginal G/inf temporal G	51	−42	36	3.77
MwoA > HC	No brain region above the threshold
Resting-state functional connectivity changes before and after acupuncture treatment
Contrast	Voxels	Brain region	MNI (x, y, z)			z
Post > Pre	1693	L MPFC/rACC	−6	66	3	4.62
		L VLPFC	−3	54	−24	4.57
		L ventral striatum	−12	18	−9	3.55
		L DLPFC	−33	−63	0	4.1
		R MPFC/rACC	9	69	6	4.2
		R ventral striatum	15	15	−12	4.16
		R DLPFC	18	69	−3	3.96
		R sup temporal G	36	6	−33	3.87
	207	L cerebellum	−42	−63	−48	4.32
	294	L mid/inf occipital G	−27	−99	−9	4.3
	305	R cerebellum	51	−69	−39	4.36
	347	R mid/inf occipital G	21	−102	12	4.51
Pre > Post	No brain region above the threshold

HC: healthy control; G: gyrus; DLPFC: dorsolateral prefrontal cortex; Inf: inferior; L: left side; MIN: Montreal Neurological Institute coordinate space; MPFC: medial prefrontal cortex; MwoA: migraine without aura; Mid: middle; R: right side; rACC: rostral anterior cingulate cortex; rFPN: right frontoparietal network; Sup: superior; VLPFC: ventrolateral prefrontal cortex.

We also found increased precuneus rs-fc with the VLPFC, DLPFC, and ventral striatum/nucleus accumbens. A previous study (62) showed that pain relief could produce negative reinforcement through activation of the mesocorticolimbic reward-valuation circuitry. Lee and colleagues found that acupuncture needles associated with a treatment context could produce greater fMRI signal increases at the ventral striatum as compared to pure needle stimulation that is associated with a treatment context (63). A more recent human fMRI–positron emission tomography study (64) found increased endogenous opioid release at the nucleus accumbens during pressure pain. Taken together, our results suggest that the cognitive control network, reward system, and descending pain modulatory network may all be involved in the modulating process of acupuncture treatment.

In summary, we believe that our findings suggest the existence of top-down coping/adaptation modulation in MwoA patients during migraine-free phases. Specifically, the brains of migraineurs have formed a reduced rFPN–precuneus connectivity mechanism in order to prepare for/cope with the potential pain attack, which may reflect the application of a top-down distraction strategy. Interestingly, we also found that the non-pharmacological treatment of acupuncture can: 1) further enhance this coping mechanism by further reducing the rFPN–precuneus rs-fc; and 2) increase precuneus rs-fc with the MPFC/rACC and ventral striatum, which may indicate the involvement of the descending pain modulation system and reward system. These results together may suggest a central pathway (including the DLPFC, VLPFC, parietal lobules, precuneus, ACC/MPFC, and PAG) underlying the acupuncture treatment of chronic pain. The elucidation of this pathway/mechanism may shed new light on our understanding of the self-regulation/coping process and facilitate the development of new treatments for migraine.

There are several limitations to this study: 1) the sample size in each acupuncture treatment group is small, which prevents us from testing clinical outcome differences between different acupuncture treatment groups. In addition, the dropout rate is relatively high; however, we would like to emphasize that the reasons for dropout do not seem to be associated with treatment response. In addition, the aim of this study is to explore the neural physiopathology of migraine, so using the treatment as a mediator was intended to enable the investigation of how an effective treatment can modulate the rFPN rs-fc, rather than to test the efficacy of acupuncture itself. Nevertheless, future studies with large sample sizes are needed in order to elucidate the specific treatment effects of acupuncture treatment. 2) Although the patients were migraine-free for at least 72 hours at the time of the MRI scan, they could have been in different stages with regards to the upcoming migraine attack (65). Studies that investigate the rs-fc changes at different phases of migraine and the modulating effects of acupuncture and other treatments will further enhance our understanding of the physiopathology of migraine. 3) We do not have a second fMRI scan for the matched HCs. Thus, we cannot completely rule out the notion that the results were observed due to drift effects over time or habituation to the scanning environment. Further studies with better controls are needed in order to replicate the results of this study.

Conclusion

Migraine is associated with abnormal rFPN rs-fc during the interictal period. The reduced rFPN rs-fc with the DMN is associated with lower headache intensity, suggesting an adapting/coping cognitive mechanism in migraine patients. An effective treatment, such as acupuncture, may achieve symptom relief by strengthening this cognitive adapting/coping mechanism.

Clinical implications

Migraineurs might be associated with abnormal right frontoparietal network resting-state functional connectivity.

An effective treatment, such as acupuncture, may relieve symptoms by strengthening the cognitive adaptation/coping process.

Elucidation of the adaptation/coping mechanism may open up a new window for migraine management.

Footnotes

Acknowledgements

The authors would like to thank Wei Qin, Jinbo Sun, Jixin Liu, Minghao Dong, Qizhu Wu, and Xiaoqi Huang from Xidian University and the West China Hospital of Sichuan University for their assistance in this study. ZL, FZ and LL contributed equally to this article. Study protocol and design: FL, FZ, YL, and QG; acquisition of data: ML, LL, TG, FW, YG, MD, and JY; analysis and interpretation of data: ZL, NM, JH, FZ, and JK; and drafting of the manuscript: ZL, LL, SS, NM, and JK. All authors reviewed the manuscript.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: grants from the State Key Program for Basic Research of China (2012CB518501), the National Natural Science Foundation of China (No.81590950, No.81273154, and No.81473602), the Program for New Century Talents in the University of Ministry of Education of China (2013) and the Youth Foundation of Sichuan Province (No.2012JQ0052); JK is supported by R01AT006364 (NCCIH/NIH), R01AT008563 (NCCIH/NIH), R21AT008707 (NCCIH/NIH), and P01 AT006663 (NCCIH/NIH).

References

Stovner

Hagen

. Prevalence, burden, and cost of headache disorders. Curr Opin Neurol 2006; 19: 281–285.

Leonardi

Steiner

Scher

. The global burden of migraine: measuring disability in headache disorders with WHO’s Classification of Functioning, Disability and Health (ICF). J Headache Pain 2005; 6: 429–440.

Guitera

Munoz

Castillo

. Quality of life in chronic daily headache: a study in a general population. Neurology 2002; 58: 1062–1065.

Diener

H-C

Dodick

Goadsby

. Chronic migraine – classification, characteristics and treatment. Nat Rev Neurol 2012; 8: 162–171.

Linde K, Allais G, Brinkhaus B, et al. Acupuncture for migraine prophylaxis. Cochrane Database Syst Rev 2009; (1): CD001218.

Zheng

Witt

. Acupuncture for migraine prophylaxis: a randomized controlled trial. CMAJ 2012; 184: 401–410.

Han

. Acupuncture analgesia: areas of consensus and controversy. Pain 2011; 152: S41–S48.

Hadjikhani

Ward

Boshyan

. The missing link: enhanced functional connectivity between amygdala and visceroceptive cortex in migraine. Cephalalgia 2013; 33: 1264–1268.

Schwedt

Larson-Prior

Coalson

. Allodynia and descending pain modulation in migraine: a resting state functional connectivity analysis. Pain Med 2014; 15: 154–165.

10.

Vecchia

Pietrobon

. Migraine: a disorder of brain excitatory–inhibitory balance? Trends Neurosci 2012; 35: 507–520.

11.

Maleki

Becerra

Brawn

. Common hippocampal structural and functional changes in migraine. Brain Struct Funct 2013; 218: 903–912.

12.

Liu

Qin

Zhang

. Combining spatial and temporal information to explore function–guide action of acupuncture using fMRI. J Magn Reson Imaging 2009; 30: 41–46.

13.

Noseda

Burstein

. Migraine pathophysiology: anatomy of the trigeminovascular pathway and associated neurological symptoms, cortical spreading depression, sensitization, and modulation of pain. In: Pain 2013; 154(Suppl. 1): S44–S53.

14.

Schwedt

Chiang

C-C

Chong

. Functional MRI of migraine. Lancet Neurol 2015; 14: 81–91.

15.

Barkhof

Haller

Rombouts

SARB

. Resting-state functional MR imaging: a new window to the brain. Radiology 2014; 272: 29–49.

16.

Corbetta

Shulman

. Control of goal-directed and stimulus-driven attention in the brain. Nat Rev Neurosci 2002; 3: 201–215.

17.

Biswal

Mennes

Zuo

X-N

. Toward discovery science of human brain function. Proc Natl Acad Sci USA 2010; 107: 4734–4739.

18.

Moriarty

Finn

. Cognition and pain. Curr Opin Support Palliat Care 2014; 8: 130–136.

19.

Moriarty

McGuire

Finn

. The effect of pain on cognitive function: a review of clinical and preclinical research. Prog Neurobiol 2011; 93: 385–404.

20.

Van Damme

Crombez

Eccleston

. Coping with pain: a motivational perspective. Pain 2008; 139: 1–4.

21.

Xue

Yuan

Zhao

. Intrinsic brain network abnormalities in migraines without aura revealed in resting-state fMRI. PLoS One 2012; 7: e52927–e52927.

22.

Russo

Tessitore

Giordano

. Executive resting-state network connectivity in migraine without aura. Cephalalgia 2012; 32: 1041–1048.

23.

Lan

Zeng

. Altered periaqueductal gray resting state functional connectivity in migraine and the modulation effect of treatment. Sci Rep 2016; 6: 20298–20298.

24.

Headache Classification Subcommittee of the International Headache Society. The International Classification of Headache Disorders: 2nd edition. Cephalalgia 2004; 24(Suppl. 1): 9–160.

25.

Zheng

Witt

. Acupuncture for migraine prophylaxis: a randomized controlled trial. CMAJ 2012; 184: 401–410.

26.

Kong

Gollub

Huang

. Acupuncture de qi, from qualitative history to quantitative measurement. J Altern Complement Med 2007; 13: 1059–1070.

27.

Headache Classification Subcommittee of the International Headache Society. The International Classification of Headache Disorders: 2nd edition. Cephalalgia 2004; 24(Suppl. 1): 9–160.

28.

Seidel

Hartl

Weber

. Quality of sleep, fatigue and daytime sleepiness in migraine – a controlled study. Cephalalgia 2009; 29: 662–669.

29.

Usai

Grazzi

D’Amico

. Psychological variables in chronic migraine with medication overuse before and after inpatient withdrawal: results at 1-year follow-up. Neurol Sci 2009; 30(Suppl. 1): S125–S127.

30.

Zuo

X-N

Jiang

. Toward reliable characterization of functional homogeneity in the human brain: preprocessing, scan duration, imaging resolution and computational space. Neuroimage 2013; 65: 374–386.

31.

Zuo

Xing

. Test–retest reliabilities of resting-state FMRI measurements in human brain functional connectomics: a systems neuroscience perspective. Neurosci Biobehav Rev 2014; 45: 100–118.

32.

Saad

Gotts

Murphy

. Trouble at rest: how correlation patterns and group differences become distorted after global signal regression. Brain Connect 2012; 2: 25–32.

33.

Jenkinson

Smith

. A global optimisation method for robust affine registration of brain images. Med Image Anal 2001; 5: 143–156.

34.

Jenkinson

Bannister

Brady

. Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage 2002; 17: 825–841.

35.

Beckmann

Smith

. Probabilistic independent component analysis for functional magnetic resonance imaging. IEEE Trans Med Imaging 2004; 23: 137–152.

36.

Filippini

MacIntosh

Hough

. Distinct patterns of brain activity in young carriers of the APOE-epsilon4 allele. Proc Natl Acad Sci U S A 2009; 106: 7209–7214.

37.

Vickers

Cronin

Maschino

. Acupuncture for chronic pain: individual patient data meta-analysis. Arch Intern Med 2012; 172: 1444–1453.

38.

Vickers

Linde

. Acupuncture for chronic pain. JAMA 2014; 311: 955–956.

39.

Smith

Fox

Miller

. Correspondence of the brain’s functional architecture during activation and rest. Proc Natl Acad Sci U S A 2009; 106: 13040–13045.

40.

Leknes

Tracey

. A common neurobiology for pain and pleasure. Nat Rev Neurosci 2008; 9: 314–320.

41.

Bushnell

Ceko

Low

. Cognitive and emotional control of pain and its disruption in chronic pain. Nat Rev Neurosci 2013; 14: 502–511.

42.

Wiech

Ploner

Tracey

. Neurocognitive aspects of pain perception. Trends Cogn Sci 2008; 12: 306–313.

43.

Pietrobon

Moskowitz

. Pathophysiology of migraine. Annu Rev Physiol 2013; 75: 365–391.

44.

Margulies

Vincent

Kelly

. Precuneus shares intrinsic functional architecture in humans and monkeys. Proc Natl Acad Sci U S A 2009; 106: 20069–20074.

45.

Utevsky

Smith

Huettel

. Precuneus is a functional core of the default-mode network. J Neurosci 2014; 34: 932–940.

46.

Raichle

Snyder

. A default mode of brain function: a brief history of an evolving idea. Neuroimage 2007; 37: 1083–1090.

47.

Buckner

Andrews-Hanna

Schacter

. The brain’s default network: anatomy, function, and relevance to disease. Ann N Y Acad Sci 2008; 1124: 1–38.

48.

Mason

Norton

Van Horn

. Wandering minds: the default network and stimulus-independent thought. Science 2007; 315: 393–395.

49.

Mansour

Farmer

Baliki

. Chronic pain: the role of learning and brain plasticity. Restor Neurol Neurosci 2014; 32: 129–139.

50.

Borsook

Maleki

Becerra

. Understanding migraine through the lens of maladaptive stress responses: a model disease of allostatic load. Neuron 2012; 73: 219–234.

51.

Quevedo

Coghill

. Attentional modulation of spatial integration of pain: evidence for dynamic spatial tuning. J Neurosci 2007; 27: 11635–11640.

52.

Villemure

Bushnell

. Cognitive modulation of pain: how do attention and emotion influence pain processing? Pain 2002; 95: 195–199.

53.

Freeman

Egorova

. Distinct neural representations of placebo and nocebo effect. Neuroimage 2015; 112: 197–207.

54.

Han

. Acupuncture analgesia: areas of consensus and controversy. Pain 2011; 152: S41–S48.

55.

Kong

Kaptchuk

Polich

. Expectancy and treatment interactions: a dissociation between acupuncture analgesia and expectancy evoked placebo analgesia. Neuroimage 2009; 45: 940–949.

56.

Kong

Kaptchuk

Polich

. An fMRI study on the interaction and dissociation between expectation of pain relief and acupuncture treatment. Neuroimage 2009; 47: 1066–1076.

57.

Jensen

Kosek

Wicksell

. Cognitive behavioral therapy increases pain-evoked activation of the prefrontal cortex in patients with fibromyalgia. Pain 2012; 153: 1495–1503.

58.

Margulies

Kelly

AMC

Uddin

. Mapping the functional connectivity of anterior cingulate cortex. Neuroimage 2007; 37: 579–588.

59.

Kong

Zyloney

. Intrinsic functional connectivity of the periaqueductal gray, a resting fMRI study. Behav Brain Res 2010; 211: 215–219.

60.

Gollub

Spaetha

. Disrupted functional connectivity of the periaqueductal gray in chronic low back pain. Neuroimage Clin 2014; 6: 100–108.

61.

Egorova

Gollub

Kong

. Repeated verum but not placebo acupuncture normalizes connectivity in brain regions dysregulated in chronic pain. NeuroImage Clin 2015; 9: 430–435.

62.

Navratilova

Xie

Okun

. Pain relief produces negative reinforcement through activation of mesolimbic reward-valuation circuitry. Proc Natl Acad Sci USA 2012; 109: 20709–20713.

63.

Lee

I-S

Wallraven

Kong

. When pain is not only pain: inserting needles into the body evokes distinct reward-related brain responses in the context of a treatment. Physiol Behav 2015; 140: 148–155.

64.

Wey

Catana

Hooker

. Simultaneous fMRI–PET of the opioidergic pain system in human brain. Neuroimage 2014; 102: 275–282.

65.

Stankewitz

Aderjan

Eippert

. Trigeminal nociceptive transmission in migraineurs predicts migraine attacks. J Neurosci 2011; 31: 1937–1943.