The present study demonstrate the presence of abnormalities in resting-state EEG microstates among MwoA patients. Overall, MwoA patients exhibited divergent temporal microstate profiles compared to those in the HC group. Moreover, the MwoA patient group, relative to the HC group, displayed multiple distinct microstate transition probabilities, which primarily involved microstate classes B, C and D. These findings Here, we will discuss how these findings may help to shed light on migraine pathophysiology.
Microstate class B and the visual network (VN)
We found a significant change in this microstate class B (time coverage and occurrence) in the MwoA patient group compared to the HC group. Microstate class B has been shown to be linked to activities in the VN encompassing the bilateral lateral extrastriate visual areas (46). A clear change in this microstate class may thus represent underlying structural abnormalities in this area in MwoA patients. This argument is supported by previous anatomical imaging studies reporting anatomical alterations in the VN (e.g., increased cortical thickness) in MwoA patients (47). Consistent with the notion that brain areas showing structural abnormalities in migraineurs also show functional alterations, structural alterations in the VN may provide an explanation for the functional changes that we observe in microstate class B.
Meanwhile, such a change in microstate class B may also indicate functional alterations of the VN in MwoA patients at baseline. That is, these results reflect activity occurring default functional states that are independent of any task performance. Occurrence and time coverage of a particular microstate class have usually been interpreted to reflect the tendency of its underlying cortical and subcortical sources to be activated as well as the corresponding relative time coverage of such underlying neural activities (22). It is thus reasonable to speculate that an increase in these parameters of microstate class B at rest in MwoA patients may indicate an enhanced likelihood of neural activation of the VN in response to visual events. This speculation is indeed supported by previous functional studies in migraineurs which have converged to reveal a functional impairment in the form of visual cortex hyperexcitability in migraineurs (48–50). In this sense, our observation of increased microstate class B activity at rest may provide a potential neurobiological mechanism to explain the visual disturbances and visual hyperactivity in the visual system of MwoA patients.
It should be noted that patients in the present study all suffered from migraines without aura. The effect observed in these patients adds to a growing literature showing that visual hyperactivity in the visual system can also be found in MwoA patients (51–53). In spite of some ongoing controversy (54) and their differing clinical symptoms (55), our finding provides adds support to the viewpoint that similar pathogenic mechanisms may be shared among all migraine patients, both with visual aura (usually with coexisting visual disturbances) and without aura.
Microstate class C and the salience network (SN)
In contrast to microstate class B, we found a significant decrease in microstate class (mean duration, occurrence and time coverage) in the MwoA patient group compared to the HC group. This microstate class has been related to the salience network (SN) focusing mainly on the dorsal anterior cingulate cortex (dACC) and anterior insula (AI) (46). Decreased microstate class C may be associated with the structural abnormalities in the SN that have been reported by previous anatomical imaging studies (56). These three microstate parameters have been interpreted to reflect three complents: the average length of time a given microstate class remains stable; the tendency of its underlying cortical and subcortical sources to be activated; and the corresponding relative time coverage of such underlying neural activities (22). Thus, a decrease in these three parameters in microstate class C may suggest a functional impairment of the SN in MwoA patients at baseline. This would be consistent with previous resting-state fMRI studies showing reduced intrinsic connectivity within the SN in these patients (57).
Regarding the functional significance of the SN, several views have emerged to provide a possible explanation of its activity during resting state. A prevailing view points to the well-established role of the SN in interceptive awareness and sensory processing of salient events (58, 59). This view emphasizes its role in detecting and filtering salient stimuli and in coordinating other brain networks (e.g., central-executive network (CEN)) to guide behavior. In addition to this view, recent studies have begun to identify a specific role of this network in inhibitory control (60, 61). Despite some controversy, one line of evidence supports the notion that the SN integrates salient information that is subsequently used by the CEN, including the inferior frontal cortex (IFC), for recruiting inhibition (61, 62). This implies an indirect involvement of salience processing in inhibitory control. Therefore, we can speculate that dysfunctional SN may lead to an aberrant assignment of salience to sensory stimuli. This is then improperly, or incompletely, processed by the CEN and ultimately causes reduced involvement of the CEN in recruiting inhibition in MwoA patients. From this perspective, it is possible that the aberrant role of salience processing in inhibitory control in MwoA patients is related to a hypervigilance to salient events (e.g., ongoing pain and sensory stimuli), which are common triggers for migraine headaches. This argument is partly supported by aberrant syntax patterns in the MwoA patient group compared to the HC group observed in the present study. A reduction in transitions from microstate class C to microstate class B and A seems to implicate decreased connectivity from the SN to the primary sensory networks. Such a pattern may then lead to an increase in cortical excitability and sensory gain, as implied by our observation of increased microstate class B in association with more engagement of the VN in MwoA patients. This finding is in agreement with previous work showing decreased SN and CEN connectivity (57) and decreased SN and VN connectivity in migraineurs (63–65). More importantly, these findings further indicate that the SN may stand at a ‘crossroads’ in the network architecture of the migraine brain and consequently may represent a potential target for improving the adverse impact of headache on daily functions in sufferers (63). Our observation of negative associations in the MwoA patient group between the mean duration of microstate class C and Hit-6 scores, which measure the adverse impact of headache on social functioning, role functioning, vitality, cognitive functioning and psychological distress, appears to support this argument.
Microstate class D and the dorsal attention network (DAN)
Finally, we observed a significant increase in time coverage and occurrence of the microstate class D in the MwoA patient group compared to the HC group. In accordance with previous work (46), this microstate class is related to activities in the DAN including the dorsal areas of the frontal and parietal cortex. Our finding thus implies a functional impairment in this network among MwoA patients. Increases in these two parameters for microstate class D at baseline seems to suggest a potential hyperexcitability of the DAN to incoming sensory stimuli. This is consistent with recent brain imaging studies showing an increase in neural responses to both attended and unattended stimuli in the key regions of the DAN (66, 67). Furthermore, the functional significance of the DAN has been argued to reflect reflexive aspects of attention, such as switching and reorientation of attention to relevant information (68). From this, we can speculate that migraineurs may exhibit an exaggerated pattern of reflexive orienting responses to incoming sensory stimuli. This argument is indeed supported by previous work showing heightened reflexive visual-spatial orienting to attended and nonattended events (69–71). In this sense, the change in microstate class D at baseline that we observed in MwoA patients would thus be associated with alterations in top-down and/or bottom-up attention during task performance. Such an interpretation receives support from our observation of aberrant syntax patterns in the MwoA patient group compared to the HC group. Here, we observed increased transitions from microstate class D to microstate class B, a finding that is also consistent with previous resting-state studies showing an increased functional connectivity between the DAN and VN (64, 72). Thus, it is possible that such atypical syntax patterns in the MwoA patient group at baseline may provide a potential neurobiological explanation for the enhanced attentional focus toward visual events described above.
Despite the relatively large sample size used in the present study, we should take several potential limitations into account. First, only MwoA patients were involved in the present study. Thus, it remains unclear as to whether these findings can be generalized to other types of migraine groups, such as patients suffering from migraine with aura (MA) and chronic migraine patients. It would be important to address this issue in future studies mainly because different pathophysiological mechanisms have been found to play a role in these different types of headache syndromes (73, 74). Second, the present study is not capable of allowing us to identify whether deviant temporal microstate profiles found in MwoA patients represent the trait or state nature of microstate abnormalities. Taking this into account in future studies would further help to shed additional light on migraine pathophysiology.