Major depressive disorder (MDD), characterized by low mood, loss of interest, possible symptoms of physical discomfort, suicidal behaviors, and cognitive dysfunction, is one of the most common mental disorders (1), and has been the second leading cause of disability worldwide because of its significant impact on the quality of life (2). In recent years, the prevalence of MDD has increased in adolescents and young adults (3), and up to 20% of adolescents are affected by MDD, which severely affected their lives and increased the risk of suicide (4). Early diagnosis and treatment contribute to decreasing MDD severity and improving the well-being and prognosis of MDD patients. Due to lack of effective and specific indicators, the diagnosis of MDD is based on patients' clinical symptoms, which is more tricky for adolescents as the manifestation of MDD among them is less typical than among adults (5, 6). Therefore, identifying potential biomarkers of depression is very important for the diagnosis and treatment of MDD, and studies on MDD among adolescent population are necessary and urgent.

A large number of functional magnetic resonance imaging (fMRI) studies have reported changes in brain structure and function among MDD patients (7, 8), including reduced hippocampal volume (9), thinner cortex in parahippocampal-limbic and insular-limbic areas (10), reduced connectivity within the frontoparietal control system (11) and disrupted network connectivity in the default mode network (DMN) (12, 13). Although many fMRI studies have explored differences between MDD patients and healthy subjects in depth, these studies cannot dynamically assess changes of brain with time. The electroencephalograms (EEG) can compensate for this defect of fMRI because of its good time resolution and can capture rapid changes in the dynamics of neuronal network (14). Because of its low cost, non-invasive, and easy to complete, EEG has been widely used in the neurocognitive disciplines in recent years. Previous studies among patients with MDD have found higher theta power in the frontal cortex and rostral anterior cingulate compared to healthy subjects (15). Higher power of the gamma band was also found among adult MDD patients with suicide ideation (16). Researches among adolescents with MDD showed lower theta band and decreased resting-state connectivity in the frontal cortex (17). Lower left-sided alpha power was also found in adolescents with MDD, and left-sided alpha power was related to depression scores (18). Currently, all of these EEG analyses are mainly based on traditional approaches, such as powder spectral analysis (18) and resting-state connectivity (17).

Resting EEG involves a limited number of potential topographic maps, as each topographic map remains stable for a certain period of time (60–120 ms) before quickly switching to another topographic map. All these topographic map are called “EEG microstates,” and they dynamically change with time in an organized manner (19). The change in the EEG microstate indicates a change in the overall coordination of neuronal activity in the participant over time (20). Although different numbers of cluster maps have been reported in previous study (21, 22), four cluster maps, which are termed A, B, C, and D, are consistently identified in majority researches (23, 24). Studies based on EEG-fMRI technical have revealed the hemodynamic correlation of EEG microstate: microstate A may be associated with auditory network, while microstate B may be associated with the visual network, microstate C may reflect activity in the default mode network and the activity of the dorsal attention network may be related to microstate D (20, 25). Based on the dynamic changes of the four EEG microstate maps, the spontaneous brain activity can be represented by the duration, occurrence, contribution and transition possibility sequence of each EEG microstate maps (20). Capturing the temporal difference of dynamic changes of EEG microstates could be a promising method to study the spontaneous brain activity of adolescents with MDD.

In recent years, EEG microstate analysis has been used in patients with mental disorder, such as bipolar disorder (26), methamphetamine use disorder (27), schizophrenia (28), and depression (21, 22). Different psychopathological conditions showed different EEG microstate feature changes. Microstate B features were found to be related to depression scores among bipolar disorder patients (26). Studies among adults with MDD found that higher occurrence of microstate A was associated with depressive symptoms (21), and the contribution and duration of microstate D were reduced among depressed patients compared to healthy subjects and related to depression severity (22). Existing studies also found that EEG microstate feature may change with medication or the use of repetitive transcranial magnetic stimulation (rTMS). The duration, occurrence and contribution of microstate B decreased among the MDD patients after 2 weeks medication (29). These findings indicate that in-depth study of EEG microstates may provide new evidence for the diagnosis and treatment of adolescents with MDD. While to date, there has been no research on EEG microstate among MDD adolescents.

Here, we used EEG microstate as a new approach to explore the difference of spontaneous resting EEG between adolescents with MDD and HCs. The aims of the present study were to test whether there are differences in the temporal characteristics of EEG microstates between adolescents with MDD and HCs and whether there is a correlation between clinical symptoms and microstates.

To the best of our knowledge, this is the first study to explore the dynamic activity of resting-state large-scale brain networks among adolescents with MDD. Our results indicate that adolescents with MDD show alterations in sub second of the whole brain. Compared to HCs, adolescents with MDD showed abnormally increased occurrence of microstate B and decreased occurrence of microstate D. The increase in microstate B may be caused by more transitions from microstate C to microstate B. Adolescents with MDD had less transitions from “A to D” and more transition from “D to C,” which were the reasons for the decrease in microstate D among MDD subjects compared with the HCs.

Limited research has examined EEG microstate duration and/or occurrence in depressed patients. Our results showed the difference in microstate duration between adolescents with MDD and HCs was not obvious, which is inconsistent with other studies in adult MDD patients (22, 39). For example, Murphy et al.'s (22) research showed that the duration of microstate D was reduced among the adults MDD patients compared to the healthy subjects. However, our results are in line with the Damborská's results which found no difference in duration between the adults MDD patients and HCs (21). Although there was no change in the duration, the occurrence and contribution changed in adolescents with MDD. Our results showed that adolescents with MDD patients had higher contribution, occurrence of microstate B. This finding is in line with a previous research which found that the duration, occurrence, and contribution of microstate B were decreased as depressive symptoms improved among MDD adults (29). The head surface signal source displays microstate B was closely associated with right posterior alpha activity by accurate low resolution electromagnetic tomography (eloreta) (40). The adolescents with MDD showed reduced posterior alpha compared to HCs, and posterior alpha activity was related to depression symptoms, anhedonia symptoms, rumination (41). Microstate class B was significantly correlated with blood oxygenation level dependent (BOLD) changes in the striatum, extra-striatal cortex, and bilateral occipital cortex, which is related to the visual network (20, 42, 43). Microstate B was also found to be correlated to posterior temporal gyrus (44). Previous fMRI research of depressed subjects has found lower functional anterior cingulate cortex and posterior superior temporal gyrus connectivity compared to HCs (45). These regions play vital roles in integrating, collecting and processing information from the external environment and the internal body (46), and understanding emotions or feeling of other people (47). The increase of microstate B among MDD patients in the current study is in line with that MDD patients have deficits in cognitive function and depression symptoms, anhedonia symptoms. The higher occurrence of microstate B was mainly caused by the increased transition from “C to B” among MDD patients. Microstate C reflects part of the DMN, which is a task negative network (20). Among patients with MDD, overactivation of the DMN may be related to negative rumination (12). Other researchers have suggested that abnormalities in DMN connectivity are associated with deficits in emotion regulation among MDD patients (48). Therefore, the more frequent transition from “C to B” may reflect that adolescents with MDD have more frequent spontaneous rumination and emotion regulation.

Our results agree with previous studies reporting a decrease in microstate D (contribution and occurrence) among the adults MDD patients (22). Microstate D was found to be closely associated to the dorsal attention network (25). The decrease in microstate D is consistent with a large number of previous fMRI studies which showed decreased connectivity of the dorsal attention network among patients with MDD (11). The dorsal attention network is involved in internally- or externally-oriented attention. Decreases in connectivity of the dorsal attention network may predict deficits in attention among MDD patients (49). In addition, connectivity between the frontoparietal network and dorsal attention network is weaker among adults MDD patients than HCs (11). Decreased connectivity of the frontoparietal network and dorsal attention network is reported to be associated with higher levels of maladaptive rumination (50). Imbalances in these different network connections could lead to not only cognitive and executive dysfunction, but also emotional regulation dysfunction, which are characteristics of MDD (51).

The current study was the first study to explore the EEG microstate changes among adolescents with first episode MDD and HCs under resting-state. Although the participants in our study was adolescents, the results were partially consistent with previous studies among adults (22, 29). Both the adolescents and adults with MDD showed an increased (decreased) occurrence and contribution of microstate B (D). There are also some different findings among adults MDD patients. For example, Murphy et al. (22) found decreased duration of microstate D while we didn't. The differences may be explained by different methodologies, such as different clustering methods and different numbers of selected topographic maps (21, 22). Moreover, all subjects in this study are teenagers and their brains are in a stage of continuous development. It has been found in previous study that the duration of microstates increases continuously with increases in age (52). One thing that cannot be ignored is that depression itself may have effect on the brain development (53), resulting in divergence at different stages. One advantage of this study is that we assessed differences between the adolescents with first episode MDD and HC groups in the general resting condition, i.e., there is no cognitive task or emotional processing, so some task-related confounders can be eliminated (54). In addition, research conducted among adolescents may have a certain predictive ability for their mental health in adulthood.

There are several limitations of this study that should be noted. First, the EEG microstate depends on source modeling technology. However, the source modeling is based on the poor spatial resolution of EEG. Secondly, the patients with MDD included in the present study are adolescents. As the first depressive episode in early adolescence may be a manifestation of a later diagnosis of bipolar disorder, some of the patients in this study may be diagnosed with bipolar disorder in later life. Thirdly, the present study only included 35 patients and 35 HCs. A larger number of subjects should be included in future studies. Lastly, although there was no gender difference between the two groups, it should be pointed out that the gender of subjects was not completely matched between groups. There are differences in EEG microstates between males and females in terms of the duration and occurrence of specific microstates (52). Therefore, future research should include more subjects to study potential gender effects.

The results of this study supported changes in microstate B and D of adolescents with MDD compared to HCs. It provided new insights into dynamic changes in resting-state EEGs of MDD adolescents, and provides some evidence for further exploration of biomarkers and early diagnosis of MDD among adolescents.

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

The studies involving human participants were reviewed and approved by National Clinical Center Medical Ethics Committee of the Second Xiangya Hospital, Central South University. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

YQH, QTY, TYY, YRZ, XPG, CXH, and XRL contributed to conception and design of the study. YQH, QTY, TYY, YRZ, KZ, XYJ, and SXW participated in data collection and investigation and evaluation of the study. YQH, QTY, and XLC performed the statistical analysis. YQH and XLC wrote the first draft of the manuscript. XLC and XRL contributed to critical revision. All authors contributed to manuscript revision, read, and approved the submitted version.

This study was supported by the National Key Research and Development Program of China (No. 2017YFC1309904), Hunan Provincial Innovation Foundation for Post-graduates (No. CX2019159).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.