Effects of Obstructive Sleep Apnea on Cognition: An Analysis of Electroencephalography Microstates and Alpha Peak Frequency

Abstract

Background: Respiration-related sleep disorders such as obstructive sleep apnea (OSA) are risk factors for mild cognitive impairment and age-related neurocognitive deficits. Nearly 60% of patients with OSA suffer from a variety of impaired cognitive functions, including attention, working memory, and episodic memory, and are susceptible to mood disorders such as depression and anxiety. However, it remains largely unexplored regarding the features of the electroencephalography (EEG) microstates of these patients and the potential association between the EEG microstates and the cognitive impairments.

Method: In this study, we performed polysomnography (PSG) on 70 patients who were recruited from the Department of Neurology and subsequently categorized into the OSA group (44 patients) and the non-OSA group (26 patients). We collected and analyzed their demographic information, blood and body fluid specimens, EEG alpha peak frequency, and EEG microstates.

Result: Compared with the non-OSA group, the OSA group reported more cases of comorbid anxiety symptoms (58.5% vs. 29%), and scored significantly lower in the word classification test. Of note, compared with the non-OSA counterparts, the OSA patients displayed markedly more microstate A, a significantly higher rate of B to A microstate conversion, and drastically lower Delta and Theta power at the F7-AV position.

Conclusion: These findings confirm that OSA may affect sleep quality and cognitive function by influencing EEG activity and microstates, which highlights a potential EEG network mechanism for OSA-induced cognitive impairment.

Keywords: EEG microstates; biomarkers; cognitive impairment; obstructive sleep apnea syndrome (OSA).

FIGURE 1

Flowchart of the admission process.

FIGURE 2

Correlation between OSA and EEG microstates. The lower edge of the box plot: First quartile (Q1, 25% quantile); the line in the middle of the box: Median (Q2, 50% quantile); the upper edge of the box: Third quartile (Q3, 75% quantile). As indicated, the upper whisker line extended to Q3 + 1.5 × IQR (interquartile spacing) or the maximum of the dataset (whichever is smaller); the lower whisker line extended to Q1 − 1.5 × IQR or the minimum value of the dataset (whichever is greater). Data points beyond the whisker line were considered outliers and indicated by a separate point. Significance of * and *** indicate the level of statistical significance. A single asterisk (*) represents p < 0.05, and three asterisks (***) represent p < 0.001.

FIGURE 3

Schematic diagram of each lead of the EEG.

FIGURE 4

Correlation between OSA and EEG alpha peak frequency. The lower edge of the box plot: First quartile (Q1, 25% quantile); the line in the middle of the box: Median (Q2, 50% quantile); the upper edge of the box: Third quartile (Q3, 75% quantile). As indicated, the upper whisker line extended to Q3 + 1.5 × IQR (interquartile spacing) or the maximum of the dataset (whichever is smaller); the lower whisker line extended to Q1 − 1.5 × IQR or the minimum value of the dataset (whichever is greater). Data points beyond the whisker line were considered outliers and indicated by a separate point. Significance of * indicate the level of statistical significance. A single asterisk (*) represents p < 0.05.

FIGURE 5

Correlation analysis of cognition‐sleep parameters. The color shades indicate the strength of the correlations, with darker colors signifying stronger correlations. The correlation coefficients in the graph range from −1 to 1, where 1 indicates a perfectly positive correlation and‐1 indicates a perfectly negative correlation. *p < 0.05, **p < 0.01.

FIGURE 6

Correlation analysis of cognition and EEG microstates. As indicated, the double letter formation signifies the direction of the conversion, e.g., DC standing for the transition from state D to state C. *p < 0.05.