Dynamics of sleep-wake cyclicity at night across the human lifespan.

Studies in adult mammals (rats, cats, mice, and humans) have revealed a surprising regularity in the duration of sleep and wake bouts. In particular, wake bout durations exhibit a power-law distribution whereas sleep bout durations exhibit an exponential distribution. Moreover, in rodents, sleep bouts exhibit an exponential distribution at all ages examined, whereas wake bout durations exhibit exponential distributions early in ontogeny with a clear power-law emerging only at the older ages. Thus, the data examined thus far suggests a similar developmental trajectory for a wide range of mammals which in turn may offer a novel metric to directly compare human and animal sleep-wake data. Therefore, we tested the generalizability of these findings by examining the distributions of sleep and wake bouts during the night in a healthy human sample - from premature infants to 70-year-olds. We find that sleep bouts elongate over the first years. At the same time wake bouts shorten but elongate again with increasing age. Moreover, sleep bout durations exhibit exponential distributions at all ages tested, except for the youngest (premature infants). Wake bouts exhibit a power-law distribution - but only during a restricted time window during adulthood. We conclude that the developmental trajectory of human sleep-wake cycles does not map well onto those of rodents; however, the method of characterizing sleep-wake cycles, using bout distribution, holds great promise for classifying sleep, its disorders, and tracking its developmental milestones across the lifespan in humans.