Search

Forty-six normal newborn infants were studied at the age of 6 to 7 days. Each infant was at first breast fed and at the next session was given a bottle feeding of the same amount of expressed breast milk on the same day. The neonatal behavior was assessed over two three-hour periods following each feeding, during which continuous polygraphic recordings and observations were made. Our study consisted of four steps: an expressed breast milk bottle feeding (step 1), an equal holding time period (step 2), an equal sucking time period (step 3), and an equal holding and sucking time period using two kinds of bottles (step 4). Infants who were breast fed had a longer period of quiet sleep and slow heart rate in active sleep compared with those who were bottle fed. In step 4, there were no significant differences in the amount of quiet sleep and low heart rate in active sleep between breast and bottle feeding. From our results, each component of breast feeding alone such as the breast milk itself, the holding time, or the time for sucking does not have any specific positive influence on neonatal behavior. Rather, the study suggests that the total mother-infant interaction during breast feeding results in positive influences on neonatal behavior.

The effect of nonnutritive sucking bursts (NNSBs) on respiratory frequency was sequentially evaluated in 12 healthy preterm infants. Studies were performed during active sleep in infants between 32 and 37 weeks postconceptional age. The duration of NNSBs was classified as follows: (a) less than 3 s, (b) between 3 and 6 s, and (c) greater than 6 s. NNSBs of each duration were found at all ages, although NNSBs > 6 s were the least frequently observed. Respiratory rate increased significantly during NNSBs of less than 3 s, and also in those of 3–6 s duration. There was no significant effect on respiratory frequency of sucking bursts longer than 6 s. Our results document an early interaction between these two motor rhythms, which is influenced by the length of the NNSB. Moreover, they indicate that the decrease in respiratory frequency reported in a similar group of preterm infants during feeding should not be interpreted as an effect of sucking per se.

Within the sleep of adults and infants there are cyclic fluctuations between quiet and active sleep. These fluctuations may also persist during wakefulness as rest-activity cycles but are less readily detected. They constitute a fundamental biological rhythm on which other daily rhythms are superimposed. In adults the rest-activity cycle is 90 minutes in duration. The quiet-active sleep cycles of term, 3-, and 8-month-old infants were determined by polygraphic recording of eye and body movements, respirations, and electroencephalogram. The cycle length at term was 47 minutes and 49 and 50 minutes at 3 and 8 months. The increase in cycle length with maturation was not significant, but there was a significant change in the proportion of quiet to active sleep within a cycle. At term they were equal, while at 8 months quiet sleep was twice as long as active sleep. Quiet sleep is a highly controlled state requiring complex feedback mechanisms. The increasing proportion of quiet sleep may be a significant measure of normal brain development.

The relationship of state to HR and behavioral responding to white noise stimulation was investigated in 2-week old infants. Results indicated that state is a potent determinant of various behavioral and HR phenomena. State was related to prestimulus HR levels. Cardiac responses elicited in active sleep were larger than those elicited in quite awake, even after prestimulus effects on responding were partialled out. Infants who were asleep throughout testing (group I) showed rapid behavioral and large HR response decrement. Infants whose state changed from active sleep to awake (group II), from quiet awake to asleep (group III), and from quiet awake to crying (group IV), showed significantly slower behavioral response decrement than those in group I, but behavioral response decrement was nevertheless evident in all groups. Cardiac response over trials also was examined in these 4 groups.

The behavioral, EEG, and auditory responses have been studied during the maturation of prematurely born infants. It was found that: (1) The less mature the infant, the less time spent in a quiet sleep state. As the infants matured, periods of quiet sleep occurred for longer durations and interrupted the active sleep phase more frequently. (2) The EEG pattern of young infants (23–26 weeks estimated conceptional age [ECA]) consisted almost entirely of an “intermittent” pattern. As maturation progressed, this “intermittent” pattern became more closely associated with periods of quiet sleep. The continuous pattern appearing at approximately 29 to 31 weeks (ECA) was increasingly associated with an active sleep phase; by 34 to 35 weeks (ECA), the correspondence between EEG pattern and sleep state was more constant. (3) All infants regardless of conceptional age responded to the auditory stimulus from all recorded areas of the scalp. A sequence of wave patterns of the auditory evoked response was noted as a function of maturation. The young infants had a large, negative, long-latency (180–270 msec) wave diffusely over the scalp. As maturation progressed, a series of negative-positive-negative components was recognized. The evoked response patterns changed wave latency-amplitude relation as a function of both age and scalp topography. (4) The latency of the evoked response waves N1 and P2 were consistently greater during the intermittent, compared to the continuous EEG pattern in infants 34 to 40 weeks age post conception. (5) A small, (early ∼ 15 msec) auditory evoked response was recorded from the ear electrode and is probably of muscle origin. This early muscle response was augmented when the baby was awake and crying, intermediate in amplitude during an active sleep period, and barely detectable or absent during quiet sleep.

The present study investigated the applicability of the Law of Initial Value (LIV) for heart rate (HR) responses evoked by a nociceptive stimulus in neonatal sleep states. Ten infants were tested during Quiet and Active Sleep. In concordance with the LIV there was a significant inverse correlation for all Ss in Active Sleep between HR change and prestimulus HR values; the average correlation was -.76, the average slope -.48. In Quiet Sleep, which had a small prestimulus HR range, this relationship was significant only for some Ss. Equating for prestimulus HR values, no significant difference existed in the magnitude of average HR response in Active Sleep vs Quiet Sleep. It was concluded that for nociceptive stimuli the initial HR value is a better predictor than the behavioral sleep state of the magnitude of HR response. It is suggested that the LIV depends largely on the modality and strength of the stimulus and the width of the prestimulus HR range.

A simple visual observation system, supplemented by measurement of skin potential, was devised for developmental studies of sleep cycles in settings where multiple electrode placement is not practicable. The findings replicated essential features of quiet and active sleep cycles which had been reported previously to exist against the background of decreasing level of physiological arousal, as sleep proceeds. Twelve newborns showed approximately one-half of their inter-feeding sleeping time in the rapid eye movement stage of sleep. Skin potential rapidly declined from the waking level, continued to decrease in level throughout sleep, increased in variability during REM sleep, and increased in level at the second waking period.

The effect of sleep on the visual evoked potentials (VEPs) in neonates was investigated in 7 preterm infants. Polygraphic monitoring including EEG, EOG, ECG, respirogram and submental EMG for the purpose of sleep staging was carried out on all infants simultaneously with VEP testing. Awake-sleep stages were classified into 4 states: awake, transitional or atypical, quiet sleep and active sleep. VEPs were recorded from Oz, referenced to Fz, in response to binocular stimulation with light-emitting diode goggles. Polygraphic and EEG date were analyzed separately. Reproducible VEPs that were appropriate for their ages were seen in all infants in the awake state. The N300 was the most reliable component across sleep states but there was a significant decrease in amplitude with quietsleep. There were no significant differences among awake, atypical or active sleep states. The other two VEP components followed the same trends as the N300. The P200, present in the older infants, disappeared in both active and quiet sleep states; the P400 was typically variable but reliably present in the awake or atypical states. When a distinction is made between quiet sleep and other arousal states, consistent and significant differences emerge. Our results emphasize the need to test infants in the same arousal states in studies of VEPs in order to make valid comparisons of latency or amplitude changes, particularly with longitudinal or follow-up studies.

To examine the functional significance of early postnatal active sleep for the development of behavioural reactivity to auditory stimuli, rat pups were daily injected i.p. from the 7th to the 18th postnatal days with 5 mg kg-1 (6.6 mmol l-1) desipramine or 25 mg kg-1 (12.2 mmol l-1) zimeldine. Sleep-wake behaviour was recorded with a static-charge-sensitive bed (SCSB) method. Both desipramine and zimeldine suppressed the percentage of active sleep relative to the total recording time throughout the treatment period. In addition, these drugs increased the percentage of quiet state and waking. At the age of 38 days the zimeldine-treated rats showed more motor activity in the open field than the controls. At the age of 39 and 78 days all rat groups behaved similarly in the open field. Startle measures and motor activation, provoked by auditory stimulation, were determined by the SCSB method when the rats were 4 months of age. Auditory stimuli, consisting of a series of ten clicks, induced a greater number of startles as well as strong movement responses in the control rats than in the desipramine- or zimeldine-treated rats. The number of small movement responses did not differ between the rat groups. These findings indicate that early postnatal active sleep and the monoaminergic systems regulating it may be important for the normal development of neuronal circuitry associated with later reactivity to auditory stimuli.

This study investigated the prolonged effects on state behavior of theophylline administered to infants for apnea of prematurity. There were three groups: Four premature infants who had received theophylline in the preterm period, five premature infants who had not received theophylline, and twenty-eight normal fullterm infants. The Theophylline infants had been off the drug for at least one month prior to the beginning of the study. Sleep-wake states were observed in the home for seven-hour periods when all infants were the same corrected ages: two, three, four and five weeks post-term. Data from the portion of the day that the infants were alone were analysed for this study. The state organization of the Theophylline group differed significantly from those of the other groups. They exhibited more non-alert waking activity, more alert, more drowse or transition, and less active sleep than did the Non-Theophylline and Fullterm infants. The state distributions of the latter two groups did not differ. On the basis of similarities between the results of this study and of a previous animal study, it was concluded that theophylline altered the normal development of state organization in premature infants. These effects persisted long after the drug had cleared the body.

Polygraphic recordings of 20 neonates with brain malformations and/or chromosomal anomalies were compared with those obtained in 29 normal fullterm neonates. Seep assessment was made with different methods described in the literature. These methods of sleep scoring gave similar results in the normal newborn babies but discrepant results in many pathological neonates. Thus sleep was also studied by other approaches; assessment of cycling of the 5 parameters (EEG, REM, EMG, Motility, Respiration) and study of the discrepant parameters during a sustained period of Active sleep (AS) or Quiet Sleep (QS), defined by the most flexible method. Newborn babies with brain malformations appeared to be poor sleepers. Amount of wakefulness is higher than in normal neonates. Some babies were nearly insomniac. Correlations of sleep patterns and anatomical lesions were possible in five cases where brain anatomy was available. A good sleep organization has been seen in babies with normal brain-stems but a normal brain stem with a normal locus coeruleus has been found in a baby with absence of AS. Absence of sleep organization has been seen in babies with an abnormal brain stem; but this anomaly was not isolated as these babies also had lesions of the forebrain. The sleep of human neonates shows a greater liability in the coordination of the different parameters than the sleep of human adults or adult animals. It appears that sleep mechanisms at the pontine level present more plasticity than the rigid mono aminergic model would suggest.

The effects of 5-hydroxytryptophan (5-HTP), a serotonin precursor, and carbidopa, a peripheral aromatic amino acid decarboxylase inhibitor, were studied in a boy with the Lesch-Nyhan syndrome. No effects on the self-mutilating behaviors of this disorder were noted. However, sleep patterns were normalized, with a reduction in the proportionate amounts of wakefulness and an augmentation of active sleep. A sharp reduction in waking behavioral dyskinesias was also correlated to increased drug dosage.

EEG polygraph recordings of 1 to over 4 h duration were obtained during daytime sleeps weekly from birth to 11–13 weeks of age in 17 normal full-term newborns. Analysis of the recordings permitted more precise specification of the time courses of early developmental changes in EEG patterns related to the sleep cycle. The trace alternant pattern of quiet sleep was seen up to 2 weeks post term in all subjects, but in none beyond 6 weeks. Active sleep onset occurred in 80% of daytime sleeps at 1–3 weeks and decreased rapidly over the next 5 weeks, but at 8–13 weeks was still seen in 5–10% of the recordings. Rolandic sleep spindle bursts appeared in some subjects as early as 4 weeks post term and were present in all beyond 8 weeks. Based upon these 3 criterion variables, the transition from ‘perinatal’ to ‘infantile’ EEG sleep patterns started at a mean age of 30.3 days and was completed at a mean age of 46.6 days. The earliest and latest completions of the transition were at 27 and 66 days, respectively. On the average the transition took just over 2 weeks. Frontal sleep transients were seen until 3 weeks post term in all subjects, but in none beyond 7 weeks. Active sleep decreased from just over 50% of total sleep time at birth to about 20% beyond 8 weeks. Percent-time quiet sleep increased proportionately. Indeterminate sleep remained relatively constant at 5–10 percent-time. A tendency for percent-time indeterminate sleep to be elevated in the presences of minor illnesses was observed. The usefulness of these, and other, EEG data in defining mature and immature EEGs in the neonatal period and the clinical significance of EEG immaturity are discussed.

Newborn rabbits were raised in isolation from Day 1 through Day 20 except for a daily 15-min feeding with the mother. They were returned to their littermates on Day 20 and remained with them until Day 30. On Days 2, 8, 16, 20, and 30, each isolate and a littermate control (group reared) were observed for 1 hr under standarized conditions. Every 10 sec a judgment was made of the animal's behavioral state. Food intake and body weight data was also obtained. In comparison to controls, isolation-reared rabbits spent significantly less time in Wake on Days 16 and 20, they spent more time in Active Sleep than in Quiet Sleep, and they had more visits to Active Sleep. Isolates also ingested less milk, had a slower rate of weight gain, and were less efficient in utilizing the food they ingested. The state findings are discussed with respect to Roffwarg, Muzio, and Dement's (1966, Science 152, 604–619) hypothesis concerning the function of Active Sleep. In general, the isolate animal differs from its group-reared control both quantitatively and qualitatively.

Auditory evoked potentials (AEPs) were recorded in 6 normal full-term newborns throughout sleep cycles on 2 consecutive days. AEP configuration, latencies, and voltages were examined and analyzed with attention to relationships with the wakefulness-sleep cycle and intra-and intersession consistency. The following conclusions have been reached: 1. 1. Previous reports of AEP differences between active and quiet sleep were confirmed. The late components of the AEP tend to be of higher voltage and longer latency during quiet sleep. The AEP of wakefulness tends to resemble that of active sleep. 2. 2. The AEP consists of 7 components. Component P2 (latency approximately 250 msec) is far the most consistently detected. Its latency and voltage are reasonably stable both within and between recording sessions, taking state into account. 3. 3. The AEP of the full-term newborn is more mature than the visual evoked potential by several criteria.

The relationship between neuronal discharge in the central nucleus of the amygdala (ACE) and timing of the respiratory cycle was assessed during quiet and active sleep and during the waking state. Of 169 neurons recorded from the ACE in intact, drug-free cats, 22% discharged phasically with the respiratory cycle during at least one sleep or waking state. The dependency between neuronal discharge and the respiratory cycle was typically strong in only one state. Forty-three percent of the respiratory-related neurons were most strongly correlated with the respiratory cycle during the waking state (AW). An additional 30% were most strongly related to the respiratory cycle during quiet sleep (QS), whereas only 11% showed the strongest dependency during rapid eye movement (REM) sleep. Half of the ACE neurons (49%) discharged at fre-quencies less than 10 spikes per second, and the most common trend in firing rate across states was one in which neurons fired more rapidly during AW and REM than during QS. No relationship between discharge rate of ACE neurons in the three states and propensity for phasic discharge with the respiratory cycle could be demonstrated.

Experimental rabbits were handled for the first 20 days of life while controls were not disturbed. At 25 and 30 days of age, each animal was observed alone for one hr and behavioral states were recorded. They were given the open-field test on Days 31–34. Handling significantly affected the organization of behavioral states and open-field performance. At 25 days handled animals spent a larger percent of sleep time in Active Sleep, and they had a greater percent of REMs than did controls. Experimental rabbits had more Active Waking and less Quiet Waking than controls at 30 days; the mean duration of time spent in Active Waking was greater than that of controls at 30 days, while the mean duration of time spent in Quiet Waking was less than controls.

Rapid Eye Movement (REM) Storms, intense REM bursts during Active Sleep (or REM Sleep), were investigated as an expression of the neural mechanisms involved in overall control of the sleep and waking states. Twenty normal full-term infants were observed in their homes for a 7-hr period when they were 2, 3, 4, and 5 weeks old. Sleep and waking states, and co-occurring behaviors of the mother, were recorded at 10-sec intervals throughout the observation day. During each 10-sec epoch of Active Sleep a rating was made of the infant's rapid eye movements, as follows: No REM, Light REMs, Moderate REMs, or intense REMs (REM Storms). REM Storm rates were related to a measure of stability in the distribution of the sleep and wake states over weeks, derived separately for time the infant was alone and for time the infant was with its mother. REM Storm rates were negatively correlated with state stability in each of these contexts. The results demonstrated a commonality between the incidence of REM Storms and overall sleep-wake state regulation.

Cyclicity within sleep was investigated in 20 fullterm infants who were monitored in the home for 24-hour periods when they were 2–5 weeks old. Using a pressure sensitive mattress in the crib, analog signals from the baby's respiration and body movements were computer scored for: Quiet Sleep, Active Sleep, Active-Quiet Transitional Sleep, Sleep-Wake Transition, and Wakefulness. Sleep episodes with at least four recurrences of Quiet Sleep were analyzed for cyclicity. Significant cyclicity of Quiet Sleep was found in 49 of the 55 sleep episodes analyzed. The mean Quiet Sleep recurrence period was 63.2 minutes. Eighteen of the 20 infants exhibited cyclicity on one or more weeks. Thus, sleep during the neonatal period is generally characterized by the presence of cyclicity. Cyclicity was not related to other parameters of the sleep cycle, but was correlated with stability of state patterning over weeks. The results provide new evidence for neurobehavioral competence of the neonate in terms of sleep organization.

The purpose of this study was to evaluate the relation between cranial blood flow (CBF) and arterial blood pressure (BP) in the human newborn. In 29 healthy term neonates simultaneous measurements of CBF using jugular venous occlusion plethysmography and BP using the neonatal Dinamap 847 were performed during active sleep (AS) and quiet sleep (QS) at least 1 h from th end of the last feed. In AS as compared to QS there was significant rise in CBF (P less than 0.001), mean arterial (P less than 0.001), systolic (P less than 0.001), diastolic (P less than 0.005) and pulse pressure (P less than 0.027). There was a significant correlation between change in the mean arterial pressure (MAP) and change in CBF (r=0.37, P less than 0.05). No correlation, however, was found between changes in CBF and changes in systolic, diastolic or pulse pressures.

Twenty-two healthy full-term infants were observed for 7 continuous hours in their homes weekly on weeks 2, 3, 4, and 5. Code recordings of their behavioral states and any mother-infant interactions were made every 10 seconds. An analysis was made of the distribution of the infant's state behaviors during the time it was left alone (mean = 279.6 minutes in 7 hours). The states were Waking Active, Quiet Alert, Fuss or Cry, Drowse or Transition, Active Sleep, and Quiet Sleep. Four state profiles were obtained for each infant, one for each week of observation; then a single quantitative measure of profile consistency over weeks was calculated. Infants with the four lowest profile consistency indices had major developmental problems between 3.5 and 30 months (p less than .001): one had aplastic anemia; a second had hypsarrhythmia, a third died of SIDS, and the fourth was hyperactive, both at home and in a testing situation. Other infants with low consistency scores showed minor developmental delays at 6 months and at 30 months. None of the infants at the upper end of the consistency continuum showed evidence of any developmental dysfunction by 30 months of age. The findings indicate that developmental inconsistency in state organization during the neonatal period is predictive of risk for later development.

The newborn actively maintains its lung volume above functional residual capacity (Olinsky et al. 1974). Inspiratory intercostal muscle tone has been suggested as an important mechanism (Lopes et al. 1981) for this and there is a reduction in functional residual capacity during active sleep when the intercostal muscles are generally inactive (Henderson-Smart and Read 1979).

Behavioral states in infant rabbits were determined from a single-channel analog recording of motility, including body movements and respiration. Behavioral states were first recorded by direct observation. Two observers simultaneously made 1-hour observations of bunnies at 7, 8, 9, 10, and 19 days of age, recording state every 10 seconds. Inter-observer reliability was assessed from epoch-by-epoch agreement on each observation. The mean percent exact agreement was 85.4. Then the animals were observed for one hour by one observer who recorded from direct observation while motility was recorded from a sensor-platform. Subsequently another person judged state from the analog recordings of motility patterns only. Validity assessment for these observations were also based on exact epoch agreement and the mean percent exact agreement was 84.0. The states recorded from motility patterns included: Waking, Sleep-Wake Transition, Active Sleep, and Quiet Sleep. Although state-related behaviors vary over this age range, the patterns of motility produced by body movements and respiration during waking and sleep states are consistently distinctive. These results suggest that the patterning of motility including motor movements and respiration is a total organismic expression of behavioral state. The procedure of motility recording permits prolonged monitoring periods and makes it possible to study the ontogeny of state organization in freely moving animals.

Spontaneous action potentials were recorded at 1 mm depth (layer IV/V) in the occipital cortex of free-moving rats between 8 and 60 days of postnatal age. Neuronal firing rates during quiet sleep (QS) increased sharply around day 11–12, parallel with an increase in the amplitude of EEG slow waves. The QS discharge pattern at all ages consisted of intermittent action potentials interspersed with short bursts. Active sleep (AS) from day 11–12 was characterized by longer lasting and more frequent bursts, and by a 2–3 × higher mean neuronal discharge rate than during QS. A peculiarity in 12-day-old rats was the presence of large fluctuations in overall firing rate continuously throughout sleep. Clomipramine completely abolished AS (for several hours) at all ages studied, during which time the cortical firing rates during sleep remained at (or lower than) the QS level prior to drug injection.

The behavior of 120 full term and 160 preterm newborn twins was assessed in several key areas: Irritability, soothability, reactivity, reinforcement value, and activity level. Infants were assessed during various situations so that aggregate, or summary, scores representing each behavioral area could be obtained. Evaluations were made during a feeding period, an active sleep period, an awake alert period during which orienting and interactional behaviors were observed, a quiet period during which reactivity to stress was observed, and any fussy periods during which irritability and soothability were assessed. Significant within-twin-pair concordance of behavior was obtained for each of the areas of behavior, suggesting the influence of constitutional variables on neonatal behavior. Analysis of these data by sex resulted in significant within-pair concordance of behavior for same-sex but not for opposite-sex twins. The actual ratings on these behaviors differentiated between full term and preterm infants. Increased behavioral deviance was observed with increasing prematurity in the areas of soothability and activity level. The results are discussed in terms of increased risk factors for preterm twins.

We studied the ontogeny and temporal organization of non-nutritive sucking during active sleep in 6 healthy preterm infants (mean birthweight 1.1 kg (range 0.8-1.3 kg) gestational age 28.6 weeks (range 26-30 w] from 30 to 35 weeks of post-menstrual (PM) age. Recognizable rhythmical sucking bursts were recorded at 28 weeks in one infant and by 31-33 weeks in the others. Results were analyzed for the periods 30-31, 32-33 and 34-35 PM weeks. The number of bursts/min increased with age, while the duration of each burst was stable (mean 4.1 s). The pause between bursts decreased. Sucking pace within bursts increased with age which resulted in an increase in the overall rate of sucking. The coefficient of variation (CV) for intersuck time within bursts and for interburst time was computed to examine the stability of the sucking rhythm. For intersuck time the CV was relatively low and constant across ages. However, for interburst time CV was relatively large across ages. Thus, there appears to be a stability of the intraburst sucking rhythm from 30 weeks of gestation, whereas the interburst rhythm is less regular. This documentation of temporal organization in sucking from 30 weeks is one of the earliest indications of an intrinsic rhythm in human behavior.

A computer-based algorithm to detect tidal breathing in long-term studies with little operator intervention is described. Recognition relies on matching the tidal signals to a set of three templates. A major consideration in the algorithm's design was its ability to detect, and, in some instances, to compensate for aberrations in the inspiratory/expiratory profiles. A test on a breath-by-breath basis of a sample of over 200 breaths in both quiet and active sleep in newborn infants showed that there was no significant difference between the algorithm-derived values for inspiratory and expiratory volumes and breath duration compared with those measured manually.

The influence of short-term aminophylline treatment on sleep behaviour was studied in six preterms infants with recurrent apnea. The incidence of apnea, respiratory pauses, and bradycardias which were closely related to the phases of active sleep, decreased during aminophylline treatment. However, the amount of active sleep remained unaffected. The mode of action of aminophylline is discussed in view of the previously proposed neurophysiological concept of apnea of prematurity.

Experimentally induced hypoxia in kittens precipitated episodes of depressed respiration and irregular cardiac function during quiet sleep, waking, and transitional states. The onset of active sleep stimulated both breathing and heart rate and decreased abnormal variability in these functions. However, hypoxia markedly reduced the proportion of active sleep. These data suggest that active sleep protects against respiratory and cardiac abnormalities in infants. Chronic hypoxemia or other factors that reduce active sleep in infants, including the normal developmental decrement in this state, may increase the risk of cardiopulmonary failure and death.

Summary: Nocturnal sleep organization was compared in normal infants and those “at risk” for sudden infant death syndrome (SIDS) (siblings and near-miss infants). Before 12 weeks of age, sleep modifications were observed in “at risk” infants. During their sleep they had a smaller percentage of intervening wakefulness with a higher amount of active sleep. Quiet and active sleep episodes had longer durations resulting in a longer sleep cycle. After 12 weeks, sleep organization tended to normalize. This fact is discussed as a possible factor for a SIDS event: a higher arousal threshold could play a critical role if homeostasy is disturbed during sleep, mainly at an age when the homeostatic control is not fully established.

Beneath the calm exterior of the sleeping mammal, dynamic forces are at play. There are mechanisms that provide for the two alternating sleep states that occur while mammals are “asleep.” During one of these states, processes are generated that result in a storm of neuronal activity; motor excitatory forces vie with motor inhibitory forces for the control of moto-neurons, and although excitation may momentarily triumph, inhibition generally reigns supreme. When the bastions of inhibition are breached by excitatory forces, a twitching and jerking of the limbs and eyes results. Because of these twitches and jerks, this state of sleep is called active sleep in infrahuman species (also paradoxical sleep, desynchronized sleep, deep sleep, etc.) and REM (rapid eye movement) sleep in humans, and the sleep stage that precedes it is called quiet sleep (or non-REM in humans) because there is little in the way of easily observable movements. These alternating episodes of quiet and active sleep comprise each sleep period.

Interest in electrophysiological studies during development has increased over the last 25 years both in animals and humans. Comparison of fetuses and neonates of different species has allowed some generalizations, showing the existence of similar trends regardless of the degree of maturation at birth (Ellingson and Rose, 1970; Ellingson, 1972; Dreyfus-Brisac, 1975).

Partial nasal obstruction was performed during a morning of quiet sleep (QS: non-REM) and active sleep (AS: REM) at ages 1 week, 2 weeks, 1, 2, 3, 4 and 6 months on 12 normal infants, 15 subsequent siblings of victims of the Sudden Infant Death Syndrome (SIDS) and 12 infants admitted for investigation of infant apnoea (‘near-miss’ SIDS). In all three groups the numbers failing to arouse after 240 s (FTA-240) in QS were significantly greater than those in AS. After 2 months of age all groups showed a decrease in the number FTA-240 in AS, whereas in QS the number did not change significantly. Subsequent siblings of SIDS had a significantly higher number FTA-240 in QS than controls. There was no significant difference in FTA-240 in QS between controls and infant apnoeas, although there was a trend for this to be higher in subsequent siblings of SIDS than infant apnoeas. It was concluded that arousal from AS is more marked than from QS, that after 2 months of age the ability to arouse from AS increases, and that in relation to SIDS, QS is the sleep state in which the infant is less able to arouse. Furthermore, subsequent siblings of SIDS differ from normal infants in their ability to arouse from QS.

SUMMARY Electroclinical features of neonatal seizures were studied in 215 neonates who developed seizures during the first three weeks of life. In 62 of them, clinical fits occurred during the actual polygraphic recording. The electro-clinical correlations were very variable, although there were some statistically significant associations between certain types of seizures and electrical discharges. The relation between seizures and sleep states was investigated in 40 infants who had more than three fits during the recording. The state in which more than 70% of seizures occurred was active sleep in 28% of the cases, quiet sleep in 13%, indeterminate sleep in 8%, and wakefulness in 8%. In 50%, seizures did not tend to occur in a given state. The authors thank Professor Suzuki for his continued interest and encouragement.

The sleep and waking states of infants are described from direct behavioral observations made in the home for a 7-hour period on weeks 2, 3, 4, and 5. States were analyzed separately for two contexts: infant alone and infant with the mother. States analyzed included Alert, Waking (Non-Alert) Activity, Fussing or Crying, Drowse or Sleep-Wake Transition, Active Sleep, Quiet Sleep, and Unclassified Sleep. Individual state profiles for each context were obtained for each week. A quantitative measure of profile consistency over weeks was calculated as an index of stability of state organization. Infants showed considerable range in state stability over weeks. State stability when alone was significantly correlated with state stability when with the mother (r=.67), even though there were differences in sleep-wake distributions and in level of stimulation in these contexts. These findings demonstrate within-individual consistency in CNS control of behavioral states, across sleep and wakefulness and throughout the diverse activities of the caretaking environment.

The cerebral blood flow (CBF) index measured by venous occlusion plethysmography in 12 normal newborn infants was found to vary significantly with sleep state. Criteria for definition of sleep state are described. The cerebral blood flow index in active sleep was 23% higher than in quiet sleep, and there were smaller, but consistent and significant, differences between intermediate and active or quiet sleep.

The neocortical EEG was quantified during sleep, while under light chloral hydrate anaesthesia, in rats ranging from 9 to 20 days of postnatal age. Continuous amplitude modulation at 0.5–1.0 min per cycle, synchronously over all frequency bands, was observed in 9- and in 10-day-old pups. By day 13, the mean amplitude had risen considerably in each of the frequency bands, and a slow fluctuation in delta-wave activity began to be discernible. Sodium glutamate injections induced EEG changes at 10 days which mimicked certain aspects of normal maturation (viz., large stereotyped potentials at irregular intervals, along with a selective increase in delta-wave activity). Natural as well as experimentally induced EEG changes were paralleled by a rise in neocortical amino acid concentrations, including glutamic acid and GABA. Delta wave amplitudes during quiet sleep increased still further between 13 and 14 days, while beta activity remained at a constant level. Between days 14 and 15 a precipitous decline in delta-wave activity was found, attributable to greater inhibition of slow waves during active sleep. No further developmental changes were noted, except for a gradual rise in the time spent in quiet sleep relative to active sleep.

Like feeding or mating behaviour, sleep is controlled by endogenous factors. It has an appetitive phase in which the animal seeks a suitable site or prepares a habitual site for sleep and assumes a sleeping posture. Finally there is the consummatory phase. However, as will be seen in the ensuing pages, it cannot be easily typified. In general, it is characterized by a number of physiological and behavioural criteria. These include a prolonged period of inactivity, circadian organization (occurring at fairly regular periods over the 24 hour period), raised thresholds of response or arousal, specific sleeping sites or postures and characteristic electro-encephalograms (Meddis, 1975). Indeed sleep may be classified into two types: active and quiet sleep.

The periodic organization of waking, quiet sleep, and active sleep was studied in control infants and siblings of victims of the Sudden Infant Death Syndrome. Spectral estimates of all-night binary state time series recorded at 1 week and 1, 2, 3, 4, and 6 months of age revealed disturbed patterns of sleep states, especially in active sleep, from as early as the first week of life. These disruptions continued until at least 6 months of age. These data support the contention that the temporal patterning of sleep state can be used as an important neurologic marker for development.

The effect of sleep state on arterial oxygen tension was determined in ten healthy term infants by means of continuous transcutaneous Po 2 monitoring. In all ten infants, Pa O 2 was significantly lower and more variable during active as compared to quiet sleep. This fall in Pa O 2 during active sleep was accompanied by asynchronous (or paradoxical) chest wall movement. These findings may be important in our understanding of the higher incidence of apnea during active sleep in both preterm and term infants.

“A most important function of time in psychology is as a matrix for imbedding the behavioral stream” (Galanter, 1984, p. 193). The temporal ordering of behaviors is of importance from the very earliest days of life, and this is apparent in the behavior of the single individual and in the behaviors of interacting individuals, such as a mother and her infant. Temporal order is considered to be expressed in two major ways: (a) in consistencies of behavioral sequences, either in an individual or between individuals; and (b) in the regularity of recurrence of specific behaviors. The first form of ordering is studied generally for the purpose of making cause-effect inferences; study of the latter form is typically motivated by an interest in the rhythmicity of behaviors such as sleep, activity, and even patterns of social interaction. In both instances, sequential analysis is the strategy for study.

Episodes of spontaneous or stimulation-related bradycardia are often observed in preterm infants. Vagal cardiac and respiratory reflexes are of major importance in this age group. The Hering-Breuer inspiratory inhibitory reflex is known to decrease with maturation. Another vagal reflex, the oculocardiac reflex, was performed during active sleep in 17 healthy preterms. The purpose of this study was to evaluate the maturation of the oculocardiac reflex in preterms and to assess the relationship betwen this cardiac vagal reflex and the Hering-Breuer inspiratory-inhibitory reflex.

In seven healthy preterm infants the transcutaneous PO2 was measured during different sleep states. The tcPO2 is lower during active sleep than during quiet sleep and shows a higher variability in active sleep. The incidence of apneic spells, which occur predominantly during active sleep, was not found to be correlated to tcPO2 levels.

The postnatal maturation of brain electrical activity and sleep-waking cycle were studied in two inbred mouse strains (C57BL and BALBc), previously differentiated in their sleep patterns at adult age. Genetic differences are evident during the first postnatal period (until day 12) in the maturation of electrical activity which is both earlier and slower in C57BL than in BALBc. On the other hand, from day 12 onwards, as soon as the sleep-waking cycle can be defined by using EEG morphology to select quiet sleep (QS) and active sleep (AS) C57BL is characterized by a higher amount of AS and a lower amount of waking (W) than BALBc, as found in juvenile and adult mice. These differences appear a little later when the recordings are performed on animals which are isolated instead of being left with the rest of the litter.

Intense rapid eye movements (REM) during sleep were investigated as a possible indication of delay in the neurodevelopment of infants. The rate of occurrence of REM storms was determined by monitoring the sleep of 15 normal, first-born infants during weeks 2 through 5 and at 3, 6, 12 months. The amount of REM within each 10-second interval of active sleep was rated on a four-point scale based on frequency and intensity of eye movements. When the babies were 12 months old, the Bayley Scales of Mental Development were administered. A significant negative correlation was found between the frequency of REM storms of 14 subjects was also studied. The negative correlation was confirmed. The findings support the suggestion that by 6 months of age REM storms express dysfunction or delay in the development of central inhibitory feedback controls for sleep organization and phasic sleep-related activities.

A densidade dos movimentos oculares rápidos (REM) é índice utilizado rotineiramente na polissonografia de adultos para avaliação da atividade fásica durante o sono REM. Analisamos aqui a densidade REM em 28 recém-nascidos de 34 a 39 semanas de idade concepcional (IC). Densidade REM foi medida por dois critérios distintos: a - Densidade REM como relação entre a duração dos movimentos oculares rápidos e a dos períodos REM. Tais índices variavam de 10,6% a 14,1% e foram mais elevados no grupo de 36-37 semanas IC. A densidade do primeiro período REM era mais elevada que a média dos demais; b - Densidade REM como percentagem dos segmentos de 10 s de sono REM nos quais ocorriam movimentos oculares rápidos. Tal critério confirmou os limites estreitos dos índices nessa faixa etária (39,6-44,6%), os quais foram mais elevados no primeiro período REM e no grupo de 36-37 semanas IC Nossos dados sugerem que também nos recém-nascidos pré-termo a densidade REM é uma característica consistente da atividade fásica e seus níveis se mantêm próximos aos da população adulta, apesar da redução ontogenética gradual na duração do sono REM. Rapid eye movement (REM) density has been helpful as a reliable index of phasic eye movements activity during REM (active) sleep in adults. We evaluated this index in 28 newborns, at 34 to 39 weeks of conceptional age (CA). The 5 hours Polysomnographie recording during sleep included electroencephalogram, electrooculogram, electrocardiogram, pneumogram, nasal thermistor for detecting airflow, and continuous oxygen saturation (tcP02) monitoring. REM density was measured by two distinct criteria: a - REM density as the duration of phasic eye movements in relation to REM periods (REMP). Such index varied from 10.6 to 14.1% and was higher in the 36-37 weeks CA group. b - REM density was calculated as a percentage of 10-second epochs of REM sleep in which phasic eye movements occurred. This criterion confirmed the narrow range of such index in this age bracket (39.9-44.6%), was higher in the first REMP and in the 36-37 weeks CA group. Our data suggest that REM density is a consistent phasic REM sleep feature in preterm newborns with levels close to the adult population, inspite of the gradual decrease of REM sleep duration with age. Higher indices were also found in the first REMP and in the 36-37 weeks CA grou by both methods.

Roffwarg et al. have proposed two complementary hypotheses to account for the finding that rapid eye movement (REM), or Active, Sleep occurs in large quantities in infancy and then drops off as the organism matures. The first assumes that the immature nervous system lacks inhibitory control and that the diminution of REM sleep reflects the maturation of the CNS. The second hypothesis postulates that REM sleep also serves a functional role, namely, to stimulate the central nervous system, thereby facilitating growth and maturation. A set of deductions, derived from the second hypothesis, was tested by analyzing the sleep-wake data of 22 normal infants. These were confirmed. The key findings were that the behavioral state of Quiet Alert correlated -0.82 with the state of Active Sleep and that this correlation was independent of maturational changes in the infants.

Serial poly graphic recordings of two to three hours duration were made in five full-term newborns with trisomy 18 and one full-term newborn with trisomy 13 syndrome. The newborns with 18 trisomy syndrome were poor sleepers with long periods of wakefulness and/or drowsiness. There were no consistent abnormalities in the sleep profile characteristic of 18 or 13 trisomy syndrome. These infants had some difficulty in organizing a stable sleep cycle especially during the immediate neonatal period. There was poor correlation between EEG patterns and states. After one to two weeks, cyclic organization of sleep showed some tendency towards normalization, although there were more trace alternant and fewer high voltage slow patterns in quiet sleep even around and after 44 weeks conceptional age. Other abnormal features of sleep often observed were an increase of indeterminate sleep, a decrease of quiet sleep, and an increase or decrease of active sleep in some records.