Changes in physiological network connectivity of body system in narcolepsy during REM sleep

Dong Yeon Son,Hyun Bin Kwon,Dong Seok Lee,Hyung Won Jin,Jong Hyeok Jeong,Jeehoon Kim,Sang Ho Choi,Heenam Yoon,Mi Hyun Lee,Yu Jin Lee,Kwang Suk Park

DOI: https://doi.org/10.1016/j.compbiomed.2021.104762

Abstract:Background: Narcolepsy is marked by pathologic symptoms including excessive daytime drowsiness and lethargy, even with sufficient nocturnal sleep. There are two types of narcolepsy: type 1 (with cataplexy) and type 2 (without cataplexy). Unlike type 1, for which hypocretin is a biomarker, type 2 narcolepsy has no adequate biomarker to identify the causality of narcoleptic phenomenon. Therefore, we aimed to establish new biomarkers for narcolepsy using the body's systemic networks. Method: Thirty participants (15 with type 2 narcolepsy, 15 healthy controls) were included. We used the time delay stability (TDS) method to examine temporal information and determine relationships among multiple signals. We quantified and analyzed the network connectivity of nine biosignals (brainwaves, cardiac and respiratory information, muscle and eye movements) during nocturnal sleep. In particular, we focused on the differences in network connectivity between groups according to sleep stages and investigated whether the differences could be potential biomarkers to classify both groups by using a support vector machine. Result: In rapid eye movement sleep, the narcolepsy group displayed more connections than the control group (narcolepsy connections: 24.47 ± 2.87, control connections: 21.34 ± 3.49; p = 0.022). The differences were observed in movement and cardiac activity. The performance of the classifier based on connectivity differences was a 0.93 for sensitivity, specificity and accuracy, respectively. Conclusion: Network connectivity with the TDS method may be used as a biomarker to identify differences in the systemic networks of patients with narcolepsy type 2 and healthy controls.