Identification of neurophysiological signatures of bipolar disorder by resting-state EEG microstate analysis

Abstract

Background

This investigation probed the neurophysiological disparities between patients with bipolar disorder (BD) and healthy controls (HC) utilizing resting-state electroencephalography (rs-EEG) microstate (MS) analysis. The study, conducted at Keio University Hospital from 2017 to 2023, sought to differentiate BD from HC for early detection.

Methods

The study included 36 BD (average age: 48.3 years) and 36 age- and sex-matched HC (average age: 47.8 years). Participants underwent a 5-min rs-EEG recording with eyes closed. MS analysis focused on the duration, occurrence, and coverage of MS (A − E) transitions. Comparisons were executed using Mann-Whitney U tests, and a logistic regression model distinguished the groups based on these MS indices.

Results

Patients with BD exhibited a shorter MS-D duration (p < 0.001) and longer MS-C (p = 0.027) and MS-E (p = 0.020) durations compared to HC. The occurrence and coverage of MS-D were significantly lower in BD (p < 0.001 for both). The logistic regression model accurately classified 84.7% of cases (χ2(3)=42.03, p < 0.001).

Limitations

Limitations include the potential influence of medication on MS dynamics and the cross-sectional design, which prevents causal conclusions. Longitudinal studies are required to comprehend the relationship between MS alterations and BD symptoms.

Conclusion

Decreased MS-D indices, along with increased MS-C and MS-E indices in BD, imply dysfunction in the frontoparietal network and impairments in the default mode and salience networks. The altered temporal dynamics of rs-EEG MSs provide a unique neurophysiological profile of BD, suggesting that MS indices could potentially serve as neurophysiological markers for BD diagnosis.