Metastability as a candidate biomarker in psychotic disorders: a narrative review

Introduction

Metastability, a concept derived from dynamical systems theory, describes the emergence and dissolution of coordinated brain activity over time. In neuroscience, it refers to the capacity of neural populations and large-scale networks to move between transient integration and relative segregation [1]. This balance allows distributed regions to cooperate while preserving autonomy for specialised functions. Psychotic disorders, particularly schizophrenia, are commonly conceptualised as disorders of dysconnectivity [2]. This review examines whether metastability can clarify temporal dysconnectivity in psychosis and support biomarker development.

Methods

A narrative review was conducted of theoretical and empirical studies on metastability, psychotic disorders, and pharmacologically altered brain states. The review focused on conceptual work distinguishing metastability from multistability, criticality, and nonspecific variability [1], resting-state functional magnetic resonance imaging (fMRI) studies linking network metastability to symptoms [2], leading-eigenvector dynamic analyses of metastability as a candidate schizophrenia biomarker [3], and reliability work assessing resting-state fMRI metastability as an individual brain-state marker [4].

Results

The literature supports metastability as a useful framework, but cautions against equating it with simple brain-state switching or generic temporal variability. Empirical measures, including variability of the Kuramoto order parameter, dynamic functional connectivity, intrinsic ignition, and leading-eigenvector-based metrics, should be interpreted as heuristic signatures rather than definitive evidence that an underlying system is metastable [1].

In schizophrenia, resting-state findings suggest symptom-relevant alterations in metastable dynamics. Lee et al. reported that connectivity and metastability measures explained 14%, 17%, 12%, and 5% of variance in positive, negative, anxious/depressive, and agitation/disorganisation symptoms, respectively; salience-network properties were particularly important, and increased salience-network metastability was associated with negative symptoms [3].

Hancock et al. found that global and local metastability were higher in people with early psychosis or schizophrenia than in healthy controls [4]. These findings suggest that psychosis may reflect deviation from an optimal metastable range, rather than a simple global increase or decrease. Excessive instability may contribute to hallucinations, aberrant salience, disorganised thought, and impaired cognitive control, whereas pathological persistence of maladaptive configurations may contribute to delusional conviction. Consistent with biomarker potential, metastability also discriminated cases from controls, reached up to 76% out-of-sample area under the curve, explained 81–92% of integration and segregation measures, and highlighted intermittent disconnectivity involving subcortical, frontal, and cerebellar regions [4].

Supporting biomarker feasibility, Sheheitli et al. found strong within-subject reliability of combined synchrony-metastability feature vectors and showed that psilocybin produced a measurable departure from baseline brain dynamics in two individuals [5]. This is not direct evidence about psychotic disorders, but suggests that altered perception and self-experience can be accompanied by quantifiable shifts in metastability.

Conclusions

Metastability offers a promising framework for linking dysconnectivity, dynamic brain-state organisation, symptom dimensions, and pharmacological modulation in psychotic disorders. In this context, it may represent a candidate biomarker of altered brain dynamics and should be further investigated in relation to diagnosis, prediction of treatment response, and assessment of relapse risk. Future studies should clarify whether altered metastability is a causal mechanism, a consequence, a compensatory process, a state marker, or an epiphenomenon of psychosis.

References

1. F. Hancock, F.E. Rosas, A.I. Luppi, M. Zhang, P.A.M. Mediano, J. Cabral, G. Deco, M.L. Kringelbach, M. Breakspear, J.A.S. Kelso, F.E. Turkheimer, Metastability demystified—The foundational past, the pragmatic present and the promising future, Nat. Rev. Neurosci. 26 (2025) 82–100.
2. K.J. Friston, The disconnection hypothesis, Schizophr. Res. 30 (1998) 115–125.
3. W.H. Lee, G.E. Doucet, E. Leibu, S. Frangou, Resting-state network connectivity and metastability predict clinical symptoms in schizophrenia, Schizophr. Res. 201 (2018) 208–216.
4. F. Hancock, F.E. Rosas, R.A. McCutcheon, J. Cabral, O. Dipasquale, F.E. Turkheimer, Metastability as a candidate neuromechanistic biomarker of schizophrenia pathology, PLoS ONE 18 (2023) e0282707.
5. H. Sheheitli, R. Hermosillo, G. Grimsrud, T. Madison, O. Miranda-Dominguez, S. Nelson, D. Fair, Z. Nahas, Metastability of resting-state BOLD fMRI as a reliable biomarker of individual brain dynamics: An interrogation of within-subject variability as a function of total acquisition time, Netw. Neurosci. (2025).