Individual consistency in behavioral states in neonates.

In this study we identified consistent characteristics of the sleep-wake state of newborn infants. Forty-one normal, full-term infants were observed in a warmer-skin temperature maintained at 36.1 degree C--located in a sound-attenuated room for 1 hr during 2 successive mid-feeding periods. The infant's state was recorded on a check-list every 10 sec using the following categories for sleep and wakefulness: Quiet Sleep A, Quiet Sleep B, Active Sleep Without REM, Active Sleep With REM, Active Sleep With Dense REM, Drowsy, Alert Inactivity, WAKING Activity, Fussing, Crying, and Indefinite State. Immediately following the completion of each observation, the 2 observers made independent judgments of the degree of organization shown by the infant, using a 6-point scale. Over the two 1-hr observations, only Quiet Sleep A, Quiet Sleep B, and Crying showed any consistency. Active Sleep as typically defined showed no consistency whatsoever. However, 2 of the 3 components of Active Sleep were significantly reliability measures for the 2 observations; in sleeping infants all 3 components of active sleep showed high reliability. These results indicate that Active Sleep may be a composite of at least 3 meaningful categories. To look at overall state organization in the neonate, including both sleep and wake states, we considered Active Sleep as 3 separate states and then combined the resulting 11 behavioral states into derived clusters such that each combination of states showed a test-retest reliability above .52. With these 5 state clusters we viewed each infant's state behaviors in terms of a profile depicting the percent of time spent in each state cluster. We found a close association between these profiles and the subjective judgments of sleep-wake state organization made by the observers. The results clearly indicate that a 1-hr observation provides reliable information on individual sleep-wake states in the newborn.