Critical transitions and self-organization in sleep dynamics

Traditionally, sleep is considered to operate according to the
classical principle of homeostasis, where physiological variables
relax back to equilibrium after perturbation. However, the complex
dynamics of sleep-stage transitions and brief arousals, which
constitute the sleep micro-architecture at time scales of seconds to
minutes, cannot be understood within this framework.

We show that arousals are intrinsic components of healthy sleep and
exhibit scale-invariant temporal organization characterized by
power-law statistics, while sleep-stage durations display exponential
behavior with characteristic time scales. The coexistence of these two
fundamentally different processes within a single regulatory mechanism
is a hallmark of non-equilibrium systems exhibiting self-organized
criticality.

We further demonstrate that this critical behavior is sustained by
structured coupling between active and quiet states, leading to
long-range temporal correlations across the sleep period. Disruptions
of this organization are associated with alterations in sleep
micro-architecture observed in sleep disorders. Together, these
findings suggest that sleep operates in a non-equilibrium critical
regime at short time scales, providing a unifying physical framework
for understanding arousals, sleep-stage transitions, and sleep
stability.