Andrew Mccarthy, Sally Loomis, Brian Eastwood, Keith A. Wafford, Raphaëlle Winsky‐Sommerer, Gary Gilmour

While several methods have been used to restrict the sleep of experimental animals, it is often unclear whether these different forms of sleep restriction have comparable effects on sleep-wake architecture or functional capacity. The present study compared four models of sleep restriction, using enforced wakefulness by rotation of cylindrical home cages over 11 h in male Wistar rats. These included an electroencephalographic-driven 'Biofeedback' method and three non-invasive methods where rotation was triggered according to a 'Constant', 'Decreasing' or random protocol based upon the 'Weibull' distribution fit to an archival Biofeedback dataset. Sleep-wake architecture was determined using polysomnography, and functional capacity was assessed immediately post-restriction with a simple response latency task, as a potential homologue of the human psychomotor vigilance task. All sleep restriction protocols resulted in sleep loss, behavioural task disengagement and rebound sleep, although no model was as effective as real-time electroencephalographic-Biofeedback. Decreasing and Weibull protocols produced greater recovery sleep than the Constant protocol, mirrored by comparably poorer simple response latency task performance. Increases in urinary corticosterone levels following Constant and Decreasing protocols suggested that stress levels may differ between protocols. Overall, these results provide insight into the value of choosing a specific sleep restriction protocol, not only from the perspective of animal welfare and the use of less invasive procedures, but also translational validity. A more considered choice of the physiological and functional effects of sleep-restriction protocols in rodents may improve correspondence with specific types of excessive daytime sleepiness in humans.