EEG signal diversity during propofol sedation: an increase in sedated but responsive, a decrease in sedated and unresponsive subjects

Transitions between wakefulness and anesthesia are accompanied by profound changes in brain functioning. A key challenge is thus to disentangle neuronal mechanisms specific to loss and recovery of consciousness, from more general effects that are not directly related to the capacity for conscious experience. Measures of neuronal diversity have been recently proposed to constitute a robust correlate of the global states of consciousness. In the present study we investigated whether EEG signal diversity is indeed related to behavioral responsiveness during propofol sedation, or rather to the general drug-related effects. To this end, we reanalyzed data collected from 20 subjects sedated with propofol. Based on the responsiveness to auditory stimuli all subjects were subdivided into two subgroups - responsive (n = 13), who remains awake throughout the experiment, and drowsy (n = 7), who becomes unresponsive during moderate sedation. Resting state EEG recorded during wakefulness and sedation was characterized by the Mean Information Gain (MIG) and Fluctuation complexity (FC) - information-theory measures estimating signal diversity or complexity, respectively. The main finding is that the drowsy group exhibited a decrease in diversity during sedation but, unexpectedly, the responsive group exhibited a robust increase in diversity (ANOVA group x state interaction: F(3) = 7.81, p 197). However, signal complexity neither differentiated the subgroups, nor decreased reliably during sedation (t-test wake vs. moderate sedation: t(19) = 2.57, p = 0.092; BF10 = 3.08). Further, we show that a change in signal diversity is negatively correlated with a delta power change (r = -0.62, p = 0.002), and positively correlated with a beta power change (r = 0.84, p < 0.001). Finally, we show that MIG behaves in a qualitatively similar manner to Lempel-Ziv - another diversity measures used in several recent studies. Overall, we revealed that propofol sedation is initially related to an increase in EEG signal diversity, and that only upon loss of responsiveness EEG diversity decreases. The qualitatively different pattern of changes in the responsive and drowsy groups makes EEG diversity a robust indirect index of responsiveness and, presumably, consciousness.