The approach at the heart of the study is EEG microstate analysis, often described as the “atoms of thought.” When a person rests quietly with their eyes closed, the scalp electrical field does not wander randomly; instead, it snaps between a small set of stable topographies, each lasting roughly 60 to 120 milliseconds before switching to another. Researchers have consistently identified four canonical microstates, labeled A through D, and decades of work have linked each one to a distinct large-scale brain network. Microstate A is associated with the auditory and language-related temporal regions, microstate B with the visual network, microstate C with the salience network that helps the brain decide what deserves attention, and microstate D with the dorsal attention and frontoparietal systems that govern orientation toward external stimuli and executive control.

Because these microstates fluctuate hundreds of times per minute, their statistical properties—how often each one occurs, how long it persists, how much of the recording it covers, and how likely it is to transition into any other—provide a compact but rich description of the brain’s intrinsic dynamics. Abnormal microstate parameters have been reported in schizophrenia, dementia, and adult depression, but the authors of the new study focused on a gap in the literature: youth major depressive disorder, a population in which suicide is a leading cause of death and in which the temporal dynamics of brain activity had been largely unexplored at this granularity.

The team recruited 177 participants aged 15 to 30 years. Ninety-five had a diagnosis of major depressive disorder and were entirely drug-naïve, meaning they had never received psychotropic medication, which removes a major confound from the analysis. Of these, 53 reported current suicidal ideation while 42 did not. Eighty-two age-comparable healthy controls completed the same protocol. Each participant underwent a standardized five-minute resting-state EEG acquisition, and the researchers applied a rigorous microstate pipeline: filtering, artifact rejection, segmentation of the global field power into brief quasi-stable topographies, and k-means clustering to extract the canonical maps, followed by fitting the maps back onto the continuous data to compute occurrence, duration, coverage, and transition probabilities between states.

Clinical characterization was thorough. Depression severity was rated with the 24-item Hamilton Rating Scale for Depression, and suicide risk was quantified with the Chinese version of the Beck Scale for Suicide Ideation. Psychosocial variables were collected alongside the neurophysiological measures, allowing the team to test whether any microstate parameter tracked the intensity of suicidal thinking rather than depression severity alone. Group differences were assessed with analysis of covariance, controlling for relevant demographic and clinical covariates, and correlations were examined within the suicidal ideation group specifically.

The results were strikingly specific. Compared with depressed youth without suicidal ideation, those with suicidal ideation showed higher occurrence, longer duration, and greater coverage of microstate A, the state associated with auditory and language processing. At the same time, they showed reduced indices for microstate B, the visual-network state, and a reduced occurrence—though not reduced duration or coverage—of microstate D, the state tied to the dorsal attention network. The pattern suggests that suicidal thinking in youth depression is not accompanied by a global disruption of brain dynamics but by a targeted reorganization involving particular networks: an amplification of the auditory-linguistic state, a dampening of the visual state, and a weakening of the externally oriented attentional state.

Perhaps the most revealing findings came from the transition probabilities, the likelihood that one microstate will be followed by another. Alterations in these probabilities were most prominent for microstates A and B, indicating that the network-level choreography—the order and rhythm with which different brain systems take turns dominating consciousness—is what most distinguishes suicidal from non-suicidal depression. Within the suicidal ideation group, the Beck Scale scores were positively correlated with the transition probability from microstate C to microstate B. In other words, the more severe the suicidal thinking, the more frequently the brain’s salience state handed control over to the visual state. Because microstate C is thought to reflect the salience network, which integrates internal emotional signals and redirects attention, and microstate B reflects visual processing, this specific transition may capture a mechanism by which distressing internal states are repeatedly rerouted into perceptual channels.

The authors interpret these alterations in the framework of large-scale resting-state networks. Microstate A maps onto auditory and language-related temporal cortex, microstate B onto the visual network, microstate C onto the salience network, and microstate D onto the dorsal attention and frontoparietal networks. The increased presence of microstate A in the suicidal group could reflect heightened engagement of verbal, self-referential, or rumination-like processing, while the decreased microstate D indices align with the well-documented impairment of executive and attentional control in suicidal states. The reduced occurrence of microstate D, without a change in its duration or coverage, hints that the attention network is recruited less often but operates normally once engaged, a subtle dissociation that coarser imaging modalities would likely miss.

What makes the study technically compelling is the combination of temporal resolution and clinical specificity. Functional MRI, which has dominated neuroimaging studies of suicidality, samples brain activity on the order of seconds and infers dynamics indirectly. EEG microstates, by contrast, capture the actual sequence of network activations as they unfold in real time, and the finding that a single transition probability correlates with the severity of suicidal ideation suggests that the risk state may be encoded not just in which networks are active but in how they hand off to one another. The drug-naïve design strengthens the inference that the observed differences reflect the illness itself rather than medication effects, and the inclusion of both a depressed non-suicidal group and healthy controls allows the researchers to separate changes associated with depression from those specifically tied to suicidal thinking.

The study is cross-sectional, so it cannot determine whether the altered microstate dynamics precede suicidal ideation, result from it, or both, and the five-minute resting-state recording captures only a snapshot of the brain’s spontaneous activity. Still, the findings open a concrete path forward. If the microstate signature of suicidal ideation can be replicated in longitudinal cohorts and refined into a reliable biomarker, clinicians could one day supplement self-report and clinical interview with an objective, inexpensive, and noninvasive measure of neural risk. For a field in which suicide prediction remains one of the most stubborn challenges in medicine, the idea that the brain’s fastest rhythms—lasting mere fractions of a second—carry information about the most consequential thoughts a young person can have is a powerful reminder that the answers to psychiatry’s hardest questions may lie in dynamics too brief for the naked eye to see but precise enough for the right instruments to capture.

Subject of Research: Resting-state EEG microstate dynamics in youth major depressive disorder with suicidal ideation

Article Title: Neural dynamics of resting-state EEG microstates in youth major depressive disorder with suicidal ideation

Article References: Bao, C., Zou, H., Chen, W., Zhang, Q., Zhou, Q., Yan, R., Yao, Z., & Lu, Q. (2026). Neural dynamics of resting-state EEG microstates in youth major depressive disorder with suicidal ideation. BMC Psychiatry. https://doi.org/10.1186/s12888-026-08582-0

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08582-0

Keywords: EEG microstates, suicidal ideation, youth depression, major depressive disorder, resting-state EEG, brain dynamics, salience network, dorsal attention network, Beck Scale for Suicide Ideation, neurophysiology, BMC Psychiatry, drug-naïve