Search

Background Simultaneous downregulation of the autonomic and central nervous system activity enables the gradual physiological state change from wakefulness to sleep onset. Dysregulation of central or autonomic arousal has been found in neurological ( Silvani et al., 2016 ) and psychiatric disorders ( Hegerl and Hensch, 2014 , Hegerl et al., 2012 , Schulz et al., 2016 , Schwabedal et al., 2016 ), often associated with dysregulated sleep-wake patterns. Aim We investigated the hypothesis that brain-autonomic co-regulation affects the attentive process in the transition from wakefulness to sleep onset. We propose that the degree to which autonomic and brain dynamics are correlated predicts the level of cortical excitability or inhibition and sleep onset behavior in this transition. To test our hypothesis, we explored electroencephalogram (EEG), electrocardiogram (ECG) and skin conductance data recorded during a 2-h resting state oddball experiment. Methods 39 healthy study participants underwent a 2-h resting EEG with eyes closed including ECG-derived heart rate (HR) and measurement of skin conductance level (SCL). The Vigilance Algorithm Leipzig (VIGALL 2.1) was used to assess brain arousal regulation based on automatic EEG-vigilance stage classification of 1-s EEG segments. These vigilance stages (=vigilance; V) were scored and cross-correlated with HR and SCL over a range of ±100 s, and a mean period of vigilance fluctuations was estimated from the frequency of maxima after low-pass filtering. Mean amplitudes of event-related potentials N100 and P200 to standard and deviant stimuli at Cz were calculated as indices of cortical excitability (N100) and cortical inhibition (P200) ( Cortoos et al., 2014 ). Results In all subjects, higher max cross-correlation coefficients (V-HR mean: r = 0.362, range: −0.069 to 0.762; SD = 0.192; V-SCL mean: r = 0.277, range: −0.299 to 0.629; SD = 0.211) were associated with longer mean periods of cortico-autonomic signals (V-HR: r = 0.462, p = .003; V-SCL: r = 0.516, p = .001). The cross-correlation of V with either HR and SCL partitioned the subjects into two groups (group no lag: n0 = 20, τ = 0; group lag: n1 = 19, τ (range) = −98 to −87) dependent on the time lag of maximal correlation. Subjects in n0, who fell asleep more often (indexed by the frequency of falling asleep: t37 = 2.49, p = .018) during the 2-h EEG compared to subjects in n1, had higher max cross-correlation coefficients (V-HR: t37 = 4.24, p = 1.43E−4; V-SCL: t37 = 5.02, p = 1.35E−5) and displayed an increased standard (t30.235 = 3.19, p = .003) and deviant P200 amplitude (t37 = 3.63, p = .001) compared to subjects in n1. No significant group differences were found for N100 amplitudes and KSS score either before or after the recording. Conclusion Healthy individuals with no time lag between cortical and autonomic signals showed a stronger brain-autonomic coupling and increased P200 amplitudes (indicating a higher level of cortical inhibition) compared to individuals with a temporal misalignment. This may explain differences in sleep-onset behavior between groups. Present results indicate possible diagnostic value in the assessment of brain-autonomic regulation, for example in disorders with sleep-related symptoms such as insomnia, dementia or depression.