Bipolar disorder is associated with enduring social, cognitive, and functional impairments even during euthymic periods. One proposed mechanism for these difficulties is alteration in self-referential processing, but prior electrophysiological studies often used emotionally valenced paradigms that may conflate self-processing with emotional bias. This study aimed to isolate self-related processing using an emotionally neutral task and to characterise the spatiotemporal and oscillatory neural signatures that distinguish self versus non-self processing in bipolar disorder.
Researchers recorded electroencephalography (EEG) from 28 participants diagnosed with bipolar disorder (type I or II) and 28 age- and sex-matched healthy control participants. Participants completed an emotionally neutral colour judgment task with two conditions: a self-related condition (preference judgment) and a non-self-related condition (similarity judgment). The task design was selected to minimise confounding emotional valence while engaging self-referential and externally guided processing streams.
Three complementary analytic approaches were used to characterise the neural response to self versus non-self processing:
Overall event-related potential amplitudes were higher in healthy controls than in participants with bipolar disorder. Controls showed greater differentiation between self and non-self conditions, with a reported condition-by-group interaction spanning 337 to 946 ms post-stimulus. This indicates that conventional amplitude measures revealed a reduced magnitude of the self-related neural response in the bipolar group.
Temporal generalisation decoding provided additional detail about the temporal persistence of the self-related neural representation. In healthy controls, decoding revealed extensive cross-temporal generalisation of the self versus non-self representation that spanned most of the trial, indicating a temporally sustained neural code for self-referential processing. In contrast, participants with bipolar disorder showed no significant cross-temporal generalisation, suggesting the representation was not temporally sustained across the trial.
The authors emphasise that reduced temporal persistence may not be detectable by amplitude-only analyses; temporal generalisation exposes a deficiency in the sustained dynamics of self-referential neural coding in bipolar disorder.
Frequency-specific decoding showed that alpha and beta bands carried information distinguishing self from non-self in both groups, but with a broader spatial and/or temporal extent in controls compared with patients. In contrast, anterior theta band activity carried self versus non-self information in the bipolar group but not in controls. These findings suggest a reconfiguration of the oscillatory substrates supporting self-referential processing in bipolar disorder, with altered contributions from theta and a relative reduction in the alpha/beta expression or extent of the self-related signal.
Exploratory correlation analyses linked decoding-derived measures to self-reported rumination and anxiety, while no association was found with manic symptoms according to the reported results. The source provides no further detail on effect sizes or statistical parameters for these exploratory correlations; those specifics were not reported in the preprint text provided.
The study concludes that the neural representation that distinguishes self-referential from externally guided processing is both smaller in amplitude and less temporally sustained in people with bipolar disorder. Reduced temporal persistence of the self-related representation is highlighted as a feature not readily apparent from conventional ERP amplitude analyses. The authors propose that these alterations in the spatiotemporal dynamics and oscillatory constituents of self-referential processing may relate to the self-related and social cognitive difficulties commonly reported in bipolar disorder.
Clinically, these findings point to altered temporal dynamics and oscillatory mechanisms underpinning self-processing in bipolar disorder, which could inform future mechanistic studies and potential targets for interventions aimed at social and cognitive function. The current report does not make clinical recommendations beyond highlighting mechanistic relevance.
The study was approved by the Joint Institutional Review Board of Taipei Medical University. The authors declared no competing interests. All data produced in the study are available upon reasonable request to the authors. Funding was declared from the National Science and Technology Council and Taipei Medical University; specific grant identifiers are reported in the source. The preprint is available under a CC-BY-NC-ND 4.0 license.