Optimization of the NICU Environment

Neurobehavioral Maturation

The expected physiologic environment for a preterm infant would be in utero, floating and moving in warm amniotic fluid, mostly in a flexed position, tasting the amniotic fluid, sensing their mother’s movements and circadian rhythm, hearing her voice and sounds from her body but being protected from loud noises, receiving rich amounts of nutrients for optimal growth, and performing gas exchange through the placenta. Full-term newborns demonstrate capabilities they have developed and learned during their fetal life: they recognize their mother's smell and voice and they are ready to crawl to the breast of the mother and start sucking from the breast during the first hour of life. From this time point onwards, the development continues in the physical and emotional closeness with the caregivers and the larger social environment, optimally by getting breast milk directly from the breast for the first months of life.

The neurobehavioral maturation of a preterm infant is based on the genetically programmed development occurring in utero, including highly active axonal and dendritic growth, myelination, synaptogenesis, and proliferation of microglia and astrocytes. This development is modified by environmental influences from a completely different environment, the neonatal intensive care unit, usually for several weeks or months. The experiences provided for the developing brain can either strengthen or weaken the development of specific neural connections, or they can modify the reading of the genetic information by epigenetic mechanisms. The time window during prematurity and early infancy is also the time for unique plasticity, allowing therapeutic interventions to compensate for structural brain injuries. Therefore, the quality of the hospital care has a large potential to affect the long-term developmental outcome of preterm infants. Medical treatments and procedures, medications and nutrition, physical and social environment, and routine infant care can be either nurturing or directly toxic to the brain. Experiences in the NICU environment should be designed to mimic the physiologic, age-appropriate in utero environment.

Depending on their gestational age at birth, preterm infants are born at different developmental stages with a large variation in the degree of immaturity of the central and autonomous nervous systems and sensory functions. Whereas a full-term newborn infant has a well-developed sleep-wake cycle; is capable of reciprocal social interaction, meaningful behaviors and responses to the stimuli from the environment; and has abilities to breastfeed, a preterm infant has an attenuated capacity to regulate his/her behavioral/sleep states and has problems processing or even tolerating a high amount of sensory information. Preterm infants have immature feeding competencies and undifferentiated and/or weak signals about their needs or stress. Their skills for social interaction are immature, especially visual interaction, emphasizing the role of the senses that develop in earlier phases of fetal life.

A preterm infant has especially poor compensatory capacities in response to stressful stimuli, which cannot be completely avoided due to the many inevitable invasive procedures in the neonatal intensive care environment. Therefore, focusing on a developmentally supportive physical and social environment is especially important during prematurity when infants are still developing their capacities to reach a balanced state, i.e., to achieve neurobehavioral organization and to be able to interact with the caregivers. Reciprocity has been shown to be a very strong factor supporting learning in fetal experiments in animal models.

There is extensive animal research and clinical observation over decades by Rene Spitz, James Robertson, Konrad Lorenz, Harry Harlow, and John Bowlby, among others (as reviewed by Sullivan et al. ), showing that newborn infants (or offspring in animal models) are biologically preprogrammed to attach to their mother/caregiver, and this attachment process is a prerequisite for normal emotional and cognitive development of the child. The mediating elements for attachment include sensory stimuli provided by the mother and hormonal and metabolic changes both in the mother and infant, making them susceptible for bidirectional bonding. Because mother/caregiver–infant bonding has been shown to be essential to neurobehavioral organization and the long-term outcome in a preterm infant, neonatal care should support parents to be physically and emotionally close to their infant and to enhance sensitive, reciprocal parent–infant interaction. The following paragraphs will describe how the NICU environment can support neurobehavioral maturation of a preterm infant and the development of parent–premature infant relationship.

Behavioral/Sleep States and NICU Environment

The behavioral states of a newborn are commonly classified into six categories, from sleep to crying. The behavioral states of a preterm infant are summarized in Table 36.1 . In preterm infants with still immature sleep states, criteria for non–rapid eye movement (NREM) and rapid eye movement (REM) sleep are not always met. Then sleep is classified as indeterminate, which can also occur as a transition between sleep states. Drowsiness can be seen as a transition between sleep and wakefulness, but it can be justified as a behavioral state in preterm infants who spend long time periods in drowsiness. Fussing can be considered as a transition from active wakefulness to crying.

Preterm infants sleep most of the time. When polysomnography is available, electroencephalography (EEG) patterns together with REMs represent the established practice to classify sleep states in preterm infants. From 28 weeks of gestation onward, the concordance between EEG and REMs begins to emerge, defining NREM and REM sleep in preterm infants. In fetuses, because EEG is not available, REMs and body movements are used instead and similarly show the concordance from 28 weeks of gestation onward. A strong biologic basis can be assumed behind the maturation of sleep states, because fetuses and neonates have identical proportions of sleep states at corresponding gestational ages ( Fig. 36.1 ).

By term, the sleep states have developed such that indeterminate sleep has disappeared and it has given space to NREM sleep, which comes ontogenically later than REM sleep. REM sleep is related to more fetal respiratory patterns such as susceptibility to apnea, suggesting that REM sleep is physiologically more immature than NREM sleep. At term, infant sleep has nearly similar proportions of NREM and REM sleep. Sleep states alternate so that one cycle of sleep states lasts for about an hour at term age and becomes longer during childhood. Fig. 36.2 shows the development of the proportions of behavioral states before and after term age.

Sleep, especially REM sleep, is important for normal brain development. Animal experiments have shown that REM sleep deprivation during early development leads to permanent deviations in behavior, to alterations in neurotransmitter responses, and to reduced brain volumes. In addition, REM sleep deprivation abolished the effects of environmental enrichment seen in nondeprived animals, suggesting that REM sleep deprivation might have detrimental effects on the later capacity for learning. In a fetus, endogenous neural cell activity of retinal ganglion cells happens during REM sleep. Cortical processing of auditory stimuli also happens during REM sleep, but not in NREM sleep, supporting the role of REM sleep in auditory learning as well. Furthermore, oxygenation in ventilated extremely preterm infants is most stable during REM and NREM sleep while indeterminate sleep or arousals are associated with more time in hypoxemia.

Based on the research knowledge on the importance of sleep on brain development of a preterm infant, all efforts should be made to develop neonatal care environments and care cultures so that they minimize sleep disturbances. A lot can be done to make the space calmer ( Table 36.2 ). Optimally, families have their own rooms so that alarms and care of other infants do not disturb the infant close by. Infant care can be delivered based on the behavioral states and cues of the infant, not based on a routine time schedule. Invasive procedures should be minimized and noninvasive monitoring prioritized (e.g., transcutaneous CO ₂ monitoring instead of blood gases). Medications such as opioids should be minimized as they affect REM sleep. We discuss below the ways to decrease sleep disturbances caused by noise and ambient light.

Awake states can be classified according to behavioral criteria in a full-term infant, but it can be difficult to define wakefulness in very preterm infants. The time spent in a quiet awake state is also short in very preterm infants, but it increases with development. The emergence of an attentive quiet awake state changes the quality of social interaction, as quiet awake state provides a window of opportunity for verbal and visual parent–infant interaction and related learning, and for feeding practices. Multiple internal and external stimuli may disturb preterm infants and may prevent them from reaching an attentive awake state. Therefore, attention is needed to control environmental stimuli and to facilitate infants’ self-soothing.

Infants at very early gestational ages cry little despite stressful and even painful intensive care. In one study, ventilator-treated very preterm infants presented crying facies for less than 1% of observation time. Crying in neurologic assessment situations increases after 31 weeks of gestation. The attenuated capacity of preterm infants to produce reactions seen in relation to pain in older infants needs to be remembered when evaluating pain responses in preterm infants. Crying increases in preterm infants after term age, reaching the peak at the average age of 6 weeks of corrected age, when the amount of crying peaks also in full-term infants. Excessive crying has been seen as a sign of poor behavioral regulation, and its role has been studied in predicting later development. Many neurobehavioral assessment tools include infant crying or irritability.

Sensory Functions and NICU Environment

The developing sensory functions form a basis for interaction between preterm infants, caregivers, and the environment. The age-specific sensory functions determine which modes of interaction are relevant at each age. The onset of sensory functions occurs in the same sequence in all mammals: the first one to occur is tactile function, followed by gustatory and olfactory, vestibular, auditory, and visual functions, suggesting a fundamental importance of this sequence for the development. However, it is also known from animal studies that normal development requires balanced multimodal stimulation.

Tactile Function and Skin-to-Skin Contact

The role of touch is central in very early developmental phases. Although preterm infants need to be carefully protected from unnecessary painful and stressful tactile stimulation, not always completely avoidable in intensive care, there is a vast literature about the benefits of parent–infant skin-to-skin contact (also called kangaroo care). Skin-to-skin contact is a natural interaction mode between the caregiver and a preterm infant when other senses are even more immature. Through skin-to-skin contact the caregiver coregulates infant physiology and behavior. It was shown in a meta-analysis of 124 controlled studies that preterm infants have a lower breathing rate and higher oxygen saturation, less pain symptoms, and better temperature and glucose balance in skin-to-skin contact compared to care in an incubator. Furthermore, the meta-analysis showed that skin-to-skin contact decreases mortality and infections and improves breastfeeding rates and the head growth of preterm infants. Furthermore, skin-to-skin care has been shown to increase vagal tone and to enhance neurobehavioral maturation in preterm infants. Consistently, very preterm infants had more optimal neurobehavioral performance at term age if the neonatal units allowed parents unlimited access and practiced daily skin-to-skin contact compared with the units with less infant-centered care.

Many hormonal mechanisms mediate the neurobehavioral effects and the psychological processes related to skin-to-skin contact. Oxytocin is one of the key hormones facilitating parent–infant relationships. Seven of eight human studies showed a strong association between oxytocin levels and the mother–infant relationship as summarized by Galbally et al. Furthermore, the fundamental biologic role of oxytocin is supported by extensive animal literature, reviewed by Rilling. Breastfeeding, skin-to-skin contact, and infants’ social signals stimulate oxytocin secretion in mothers. Oxytocin may promote early interaction by several mechanisms, such as improved face and affect recognition and increased caregiving motivation by activation of the dopaminergic system, the “reward pathway.” Maternal behavior is increased by estrogen, which increases the number of oxytocin receptors.

Consistent with reports showing that massage, a type of tactile stimulation, increases growth in preterm infants, an Italian group showed in a randomized study that infant massage 15 minutes three times a day increased blood levels of insulin-like growth factor 1 (IGF-1). In addition, the group receiving massage had accelerated maturation of visual function and EEG. In rat pups, the beneficial effects on vision were blocked by antagonizing IGF-1.

The effects of skin-to-skin care on improved growth might partially be mediated by cholecystokinin (CCK), which enhances digestion and has a calming effect. Preterm infants in skin-to-skin care had a CCK response to tube feeding while tube feeding without skin-to-skin care did not increase CCK levels. Skin-to-skin contact and breastfeeding also decrease cortisol levels in mothers. A care setting supporting parental presence in a NICU improved cortisol synchrony between the mother and the infant.

Most commonly, parent–infant skin-to-skin contact has been implemented during neonatal intensive care and practiced for a few hours daily. Although skin-to-skin contact has been shown to be safe, even in extremely low gestational age infants from the first week of life ( Fig. 36.3 ), the beginning is commonly delayed and the average daily duration of skin-to-skin contact in preterm infants remains low in most neonatal units. In a study done in 11 European neonatal units, the duration of skin-to-skin contact was longer in the neonatal units providing an opportunity for the parents to stay overnight in the neonatal unit. Comfortable, reclining chairs or adult beds at the infant bedside can be provided even without family rooms to facilitate longer periods of parent–infant skin-to-skin contact. Flacking and Dykes explored the role of space in neonatal units from the perspective of mother–infant attunement at feeding interaction. The researchers concluded that the bed or chair at bedside signaled the importance of the parents and what was expected of them. When a bed was at bedside, parents were signaled that they were expected to lie in it and that a bed was needed for parent–infant closeness. A chair at bedside conveyed a message that parents were expected to come and stay close to the infant, possibly holding him/her for as long as they could.

The value of early contact in the formation of the mother–infant bond was shown in full-term infants in the landmark study by Klaus et al. in 1972 and has been confirmed later by them and others. There is evidence that early physical contact between a preterm infant and the mother in the delivery room is associated with fewer later behavioral problems at 5 years of age. However, it remains challenging to apply continuous skin-to-skin care for preterm infants right after birth in the delivery room. Poor implementation of early skin-to-skin contact in preterm populations might be defended by safety concerns, but it is more likely that the obstacles are related rather to the care culture and attitudes, which do not sufficiently appreciate the value of mother–infant closeness. The preterm infant is optimally transported to intensive care on the parent's chest, and the mother and her preterm infant are cared for together in the same room throughout the hospital stay (couplet care). Flacking and Dykes concluded in their observational study of four neonatal units that the earlier and the longer a mother can stay with her baby in close proximity, the more she will be attuned to the signals of the baby and assume the role as the primary caregiver.

In addition to tactile stimuli, skin-to-skin contact with a parent provides the infant with other stimuli, including kinesthetic, olfactory, and auditory stimuli, and an interactive dimension associated with the parents’ immediate responses to their infant's cues. Some of these elements can be achieved by holding the infant (with clothes on). All caring and gentle touch should be encouraged in all units, even in those that have not yet developed competencies for early and prolonged skin-to-skin contact.

Circadian Rhythm and NICU Environment

Circadian rhythmicity has developed during evolution to adjust life to the rotation of earth. Specific genes have been identified that regulate circadian rhythmicity, yet it is profoundly affected by the sensory environment, especially light. In newborn infants, the shorter (about 3-hour) cycles regulated by feedings dominate and circadian rhythm is not yet established. Wakefulness starts to cluster to daytime during the second month of life, and clear rhythm following the 24-hour day is normally developed by 1 year of age.

During fetal life, circadian rhythmicity operates quite differently than during the rest of life. The fetus has a well-developed circadian rhythm by the end of the second trimester, probably influenced by multiple maternal stimuli, including physical (movement, body temperature) and hormonal (transplacental melatonin, cortisol) zeitgebers (timekeepers). In preterm infants, a circadian rhythm cannot be detected, even though the neural mechanisms to respond to a circadian light stimulus are intact by early in the third trimester. In exploring this phenomenon, one is struck by the absence in the NICU of the circadian stimuli that would be available to the fetus. Body temperature is rigidly servo controlled, touch and kinesthetic stimuli are as likely to occur in the middle of the night as during the day, and access to maternal rhythms of melatonin and cortisol, present in breast milk, is often prevented by administration of breast milk without regard for the time it was expressed or by the use of artificial formula. Even skin-to-skin contact with the mother for several hours a day may not be sufficient to impart circadian information to the preterm infant. Preterm infants may develop a circadian rhythm sooner if exposed to cycled lighting.

Sensory Functions and Pain

Very preterm infants have not yet developed clear behavioral pain responses. The robustness of pain responses (crying, arousal, and facial grimace) increases with development. Even if poorly capable of displaying clear behavioral responses, preterm infants are sensitive to pain. Blood flow changes in the contralateral somatosensory cortex have been shown to be related to a heel stick as early as 25 weeks of gestation, and the blood flow responses are stronger at lower gestational ages compared with more mature preterm infants. Recurrent pain exposure in preterm infants is suggested to delay their growth, alter their brain maturation and the performance in neurobehavioral assessments, and impair their motor and cognitive development.

As both pain and pain medications are potentially harmful for brain development, the optimal solution is to avoid both of them. The care of preterm infants has become less invasive and more gentle, which has been reflected in decreasing rates of neurologic impairments in preterm infants shown in both European and Australian register studies. Interestingly, Lester et al. showed in their study about single-family rooms that the infants cared for in a single-family room were exposed to fewer procedures.

Skin-to-skin contact in different forms has been used to alleviate pain and stress in preterm infants. Parents can effectively alleviate the stress/pain behavior of their preterm infant by holding the infant by their warm hands in a flexed position during a painful or stressful procedure.

Muscle Tone, Movements, and Reflexes

Preterm infants typically have lower muscle tone and less coordinated movements than full-term infants. A flexed position maintained most of the time in utero supports infants’ self-soothing abilities such as bringing the hand to the mouth ( Box 36.1 ). To provide similar opportunities for preterm infants, special attention in nursing care needs to be paid to supporting a flexed position in a preterm infant both during rest and during handling. Although healthy preterm infants by term can reach the same level in muscle tone, tone distribution, spontaneous movements, and reflex functions as full-term infants, preterm infants present a larger variation in their performance, especially in muscle tone. This variation makes it more difficult to define neurologic normality in preterm infants at term equivalent age than in full-term infants. Preterm infants at term tend to have less flexor limb tone, less head control, and increased hyperexcitability seen in brisk reflexes, strong grasp, startles, and tremors. More tremors are seen in preterm infants born in the lowest gestational weeks.

The flexed position of a preterm infant should be supported in their environment whether it is in an incubator or on a parent's chest, or during rest and handling. It is an open question how much movement (vestibular stimulation) the care environment should offer for a preterm infant and whether a more physiologic upright position could be feasible.