Preserved calibration of persistence based on delay-timing distribution during sleep deprivation

We frequently encounter decisions where we have to determine whether to wait for a certain reward delayed for an uncertain duration or to move on. The appropriate decision depends upon the underlying temporal distribution of the delay. With some distributions it is best to be completely persistent, whereas in others it is more appropriate to abandon waiting after a certain period of time. The current study examined whether the ability to form temporal expectations and adjust persistence accordingly is compromised by sleep deprivation. Participants performed a willingness-to-wait task either in a well-rested state or after a night of total sleep deprivation. Participants had to decide either to wait for a larger reward or to abandon waiting in favour of a smaller immediate reward. Delays were drawn from either a uniform distribution, where being persistent yields maximal returns, or from a heavy-tailed distribution, where occasional long delays render full persistence suboptimal. In spite of increased sleepiness and decreased vigilance, sleep-deprived participants were able to adjust waiting time appropriate to the experienced timing distribution. Additionally, sleep deprivation did not affect the foreperiod effect, indicating intact perception of conditional probability of temporal events and ability to adjust preparation accordingly.