A three states sleep–waking model

The mechanisms underlying the sleep-states periodicity in animals are a mystery of biology. Recent studies identified a new neuronal population activated during the slow wave sleep (SWS) in the ventral lateral preoptic area of the hypothalamus. Interactions between this neuronal population and the others populations implicated in the vigilance states (paradoxical sleep (PS) and wake (W)) dynamics are not determined. Thus, we propose here a sleep–waking theoretical model that depicts the potential interactions between the neuronal populations responsible for the three vigilance states. First, we pooled data from previous papers regarding the neuronal populations firing rate time course and characterized statistically the experimental hypnograms. Then, we constructed a nonlinear differential equations system describing the neuronal populations activity time course. A simple rule playing the firing threshold role applied to the model allows to construct a theoretical hypnogram. A random modulation of the neuronal activity, shows that theoretical hypnograms present a dynamics close to the experimental observations. Furthermore, we show that the wake promoting neurons activity can predict the next SWS episode duration.