[Distribution of electroencephalograph power density in patients with severe obstructive sleep apnea during different sleep stages].

Objective: To investigate the variation of electroencephalograph(EEG) power density during different sleep stages in OSA for understanding of the mechanisms underlying the brain dysfunction in OSA as well as its earlier diagnosis and treatment. Methods: Sixteen-channel EEGs from OSA patients and normal controls in stage wake, sleep stage 1, sleep stage 2, sleep stage 3 and rapid eye movement stage were analyzed by time-frequency analysis method. The EEG power density in different frequency bands (including δ, θ, α, σ, β and γ) was respectively compared between the 2 groups. The correlation between the variation in the EEG power and primary indices of polysomnography was further analyzed. Results: The EEG power density in δ band in stage wake [OSA: (0.82±0.13) μV(2)/Hz, Control: (0.66±0.02) μV(2)/Hz, t =4.309, P <0.05], stage 1 [OSA: (1.28±0.07) μV(2)/Hz, Control: (0.92±0.04) μV(2)/Hz, t =-3.369, P <0.05] and stage 3 [OSA: (2.74±0.22) μV(2)/Hz, Control: (2.04±0.07) μV(2)/Hz, t =-2.669, P <0.05] was significantly higher in OSA, compared with that in the control. Statistical analysis showed that the EEG power density was significantly higher in frontal and central regions in stage wake [frontal: OSA: (0.90±0.02) μV(2)/Hz, Control: (0.66±0.02) μV(2)/Hz, t =8.539, P <0.01; central: OSA: (1.15±0.06) μV(2)/Hz, Control: (0.72±0.02) μV(2)/Hz, t =6.669, P <0.01] and stage 1 [frontal: OSA: (1.23±0.03) μV(2)/Hz, Control: (0.99±0.03) μV(2)/Hz, t =5.983, P <0.01; central: OSA: (1.52±0.05) μV(2)/Hz, Control: (1.14±0.04) μV(2)/Hz, t =5.714, P <0.01], as well as central region in stage 3 [OSA: (3.24±0.17) μV(2)/Hz, Control: (2.71±0.08) μV(2)/Hz, t =2.707, P <0.05]. The correlation analysis showed that the power density in central region in stage 1 and stage 3 was positively correlated with arousal index ( r =0.877 in stage 1, 0.656 in stage 3), implying that sleep fragmentation was closely related to the variation of EEG power density during nocturnal sleep in OSA. Conclusions: The feature stages for OSA are stage wake, stage 1 and stage 3. The EEG power density in OSA (δ band) was significantly higher than that in the control. The EEG power density in OSA and the control shows differences in frontal and central regions in stage wake and stage 1, as well as in central region in stage 3. The results indicate that low-frequency EEG power density giving priority to frontal area and central area has improved in severe OSA, which may be related to the neurologic deficits in corresponding brain areas.