Light does not merely entrain the biological clock to the 24-hour cycle: it also exerts direct effects, independent of the circadian system, on sleep, wakefulness, and behavior. The classical model of sleep regulation rested on two interacting processes, one driven by the circadian clock and the other homeostatic in nature—that is, linked to the accumulation of sleep pressure across the waking period. Recent data obtained in rodents have led to the proposal of a third process, namely that of direct photic influence. This work aims to precisely characterize these direct effects of light and darkness as a function of circadian time, and to determine whether they depend on the duration of the exposures.
To this end, mice were subjected for 24 hours to ultradian light-dark cycles, that is, a rapid alternation of light and dark phases. Two protocols were compared: a T2 cycle alternating 1 hour of light and 1 hour of darkness, and a T7 cycle alternating 3.5 hours of each. Cortical (ECoG) and muscular (EMG) activity was recorded in order to classify vigilance states—wakefulness, slow-wave sleep (NREM), and paradoxical sleep (REM)—and to analyze sleep quality through the spectral powers of the delta, theta, and gamma bands. Values were systematically compared with a reference condition under a normal 12-hour/12-hour cycle (T24).
The results show that exposure to light clearly promotes slow-wave sleep, with a lesser effect on paradoxical sleep, reduces the amount and quality of wakefulness, and alters sleep depth. These effects are strongly modulated by circadian time: they appear mainly during the animals' subjective night, particularly in the early night (between CT12 and CT18 under the T2 cycle). The duration of the pulses proves to be decisive for the kinetics of these modulations. Under one-hour pulses, the effects are maintained throughout the exposure, the time being too short for a homeostatic pressure to build up. In contrast, under 3.5-hour pulses, the effects of light gradually wane after one hour, and the changes induced by darkness even reverse in the second half of the phase, indicating that the homeostatic process then takes precedence over the direct influence of light.
The authors emphasize that light exposure mistimed relative to circadian time—light during the subjective night or darkness during the day—alters the distribution of waking and sleep, vigilance, and sleep depth, even in the absence of sleep deprivation. They note that these data have immediate applications in humans, notably for adapting the lighting of shift workers and, more broadly, optimizing societal lighting.