Sleep periods exhibit numerous intermittent transitions among sleep stages and short awakenings, with fluctuations within sleep stages that may trigger micro-states and arousals. Despite the established association between dominant brain rhythms and emergent sleep stages, origin and functions of sleep-arousals and sleep-stage transitions remain poorly understood. Empirical observations of intrinsic fluctuations in rhythmic cortical activity, and the corresponding temporal structure of intermittent transitions in sleep micro-architecture, raise the hypothesis that non-equilibrium critical dynamics may underlie sleep regulation at short time scales, in co-existence with the well-established homeostatic behavior at larger time scales. In this talk, I will discuss recent results on the dynamics of dominant cortical rhythms across the sleep-wake cycle that support such hypothesis (Lombardi et al. J. Neurosci. 2020; Wang et al. Plos. Comp. Biol. 2020 ). I will focus on cortical theta and delta rhythms in rats, which are associated with arousals/wakefulness and sleep respectively. I will show that intermittent bursts in theta and delta rhythms exhibit a complex temporal organization: Theta-burst durations follow a power-law distribution, whereas delta-burst durations follow an exponential-like behavior. Such features are typical of non-equilibrium systems self-tuning at criticality, where the active phase is characterized by bursts with power-law distributed sizes and durations, while quiescent periods (inactive phase) are exponentially distributed. By interpreting theta-bursts as active phases and delta-bursts as inactive phases of the cortical activity in the sleep-wake cycle, I will then draw a parallel with other non-equilibrium phenomena at criticality, and demonstrate that theta-bursts exhibit a peculiar organization in time described by a single scaling function (Gamma distribution) and closely reminiscent of earthquake dynamics. Overall, such results constitute a fingerprint of critical dynamics underlying the complex temporal structure of intermittent sleep-stage transitions at the behavioral level, and complement previous observations of critical behavior at the neuronal level (we find similar scaling exponent).
