The Vulnerable Neonate and the Neonatal Intensive Care Environment
Penny Glass
Environmental factors in the neonatal intensive care unit (NICU) have major implications for the care of the sick newborn infant. Advances in medical technology during the last 3 decades are credited with dramatic reductions in mortality, with a 50% survival rate for newborns weighing 1,500 g in 1970 to a 50% survival rate for those weighing less than 700 g by 2000. Morbidity among survivors, however, is a problem of increasing proportions. Whereas the rate of major morbidity has remained fairly stable, around 10%, this focus on major morbidity has overlooked the much larger number of children born prematurely who have learning disabilities at school age. Broad evidence implicates the environment in the NICU as a factor in neonatal morbidity. Abnormal sensory input can be a source of potentially overwhelming stress and, at a sensitive period during development, can modify the developing brain. The NICU environment, therefore, assumes a crucial role in the care of the sick newborn infant.
Preterm birth is the most common single risk factor for developmental problems in childhood, and learning disability is the most pervasive developmental problem. This is a catch-all term, but includes children of low, average, or otherwise normal intelligence who have deficits in language, visual perception, or visuomotor integration; deficiencies in attention span, hyperactivity; or social immaturity. Such children require either special services to function in a regular classroom or placement in a special class. Reports of school-age children who were of very-low-birth-weight indicate that as many as half have learning disabilities (1,2,3,4,5,6,7,8). Such deficits may originate from overt damage to the brain or from a more general disturbance in brain organization.
Throughout infancy, both behavioral and neurologic differences exist between full-term and preterm infants, even when matched for conceptional age. The latter often exhibits manifestations of altered brain organization, including disrupted sleep, difficult temperament, both hyperresponsivity and hyporesponsivity to sensory input, prolonged attention to redundant information, inattention to novel stimuli, and poor quality of motor function (9,10,11,12,13,14,15). These precursors of learning problems in school are not fully explained by either the severity of illness among the preterm infants or by later conditions in the home environment (10).
The sensory environment in the NICU is different in virtually every respect, both from the environment of a fetus in utero and from that of a full-term newborn at home. The NICU experience also contains frequent aversive procedures, excess handling, disturbance of rest, noxious oral medications, noise, and bright light. These conditions are sources of stress and anomalous sensory stimulation, both of which may affect morbidity.
The immediate effects of stress are autonomic instability, apnea/bradycardia, vasoconstriction, and decreased gastric motility. Cortisol, adrenaline, and catecholamines are secreted during stress as part of an intricate hypothalamic-pituitary-adrenocortical system (16,17). High levels of these hormones interfere with tissue healing. Noxious stimuli disrupt sleep and can have biological consequences for the neonate. Even medical complications commonly associated with prematurity per se, such as bronchopulmonary dysplasia and necrotizing enterocolitis, may be, in part, stress-related diseases (18).
Sensory input is essential during maturation. Most of the cortex is part of one of the sensory systems. Abnormal experiences, both depriving and overstimulating, can modify the developing brain. The most vulnerable period occurs during rapid brain growth and neuronal differentiation
(19,20). The timing of these events for the human fetus corresponds to 28 to 40 weeks of gestation (21). It is assumed that, for the fetus, the optimal sensory environment is experienced within the womb. One of the most striking aspects of this environment is the bidirectional contingency between mother and fetus.
(19,20). The timing of these events for the human fetus corresponds to 28 to 40 weeks of gestation (21). It is assumed that, for the fetus, the optimal sensory environment is experienced within the womb. One of the most striking aspects of this environment is the bidirectional contingency between mother and fetus.
The potential impact of the anomalous NICU environment on the vulnerable newborn infant has raised unabated concerns for more than 2 decades (22,23,24,25,26,27,28). A more optimal NICU environment might reduce iatrogenic morbidity and improve the outcome of sick neonates; however, the parameters are not yet well defined. This chapter summarizes the maturation of each sensory system during late fetal development, with particular reference to evidence for the prenatal onset of function, compares the intrauterine and NICU sensory experience, and critiques techniques of developmental intervention.
NEONATAL SENSORY SYSTEMS: DEVELOPMENT, DISORDERS, ENVIRONMENT, AND INTERVENTION
Maturation of all the sensory systems begins during the latter part of embryogenesis; however, the process is neither unitary nor fixed. Within each system some reciprocity between structure and function probably exists. To some extent, sensory input drives maturation (29). In addition, the rate of maturation of each sensory system varies, with the onset of function generally in the following order: tactile, vestibular, gustatory-olfactory, auditory, and visual (30). These sensory systems also are interrelated in a hierarchical manner—stimulation of early maturing senses (e.g., tactile, vestibular) has a positive influence on development of later-maturing ones (e.g., visual) (31). Recent research also indicates that untimely stimulation within this sequence (e.g., visual) may disrupt the normal maturational process of another sensory system (e.g., auditory) (Philbin MK. personal communication, 1998.).
This hierarchical organization and integration of sensory function is well supported and bolsters two guiding principles for developmental intervention in the NICU:
Stimulation of the senses should begin with the most mature.
The optimal form of stimulation for initial postnatal development resembles the sources naturally available to the fetus and infant—those that come from the mother.
Tactile System
The cutaneous system includes sensation of pressure, pain, and temperature. Only pressure is discussed here; pain is discussed in Chapter 57. Receptors in the skin respond to pressure and then transmit impulses to the spinal cord through the dorsal root, ascending in the posterior tract and terminating in the gray matter of the cord. At this point, connecting fibers decussate and continue in the ventral spinothalamic tract to the medulla and the thalamus, terminating in the postcentral gyrus of the cortex. Representation here is somatotopic and contralateral to the stimulated side. Increased stimulation to an area of the body or loss of a limb can alter the pattern of representation in the somatosensory cortex.
Development
Like the vestibular system, the tactile sense develops early in fetal life and is thought to play a particularly pervasive role in the early development of the organism. Receptor cells are present in the perioral region in the fetus by 8 weeks of gestation and spread to all skin and mucosal surfaces by 20 weeks. The cortical pathway is intact by 20 to 24 weeks of gestation, and some myelin is already present. Response to tactile stimulation has been observed by ultrasound as early as 8 weeks of conceptional age (32). Response to stroking in the lip region occurs first, followed by a response to stimulation of the palms. Most of the body is sensitive to touch by 15 weeks (33).
Tactile threshold is very low in the preterm infant. It is more related to postconceptional age (PCA) than to natal age but increases by term. Infants younger than 30 weeks PCA respond by an unequivocal leg withdrawal to pressure of a 0.50-g von Frey hair applied to the plantar surface of the foot compared to 1.7-g pressure by 38 weeks PCA (34). A qualitative shift occurs around 32 weeks PCA. Infants less than 32 weeks PCA respond to repeated stimulation with sensitization and a diffuse behavioral response. In contrast, infants after this age show habituation to the same stimuli.
Classic studies by Harlow and Harlow (35) demonstrated the profound importance of contact comfort for normal development. In a parallel fashion, even preterm infants will seek and maintain contact with a physical object within their incubator and even more so if the tactile source contains rhythmic stimulation (36). These findings provide strong support for intervention in the tactile modality.
Disturbances
Tactile hypersensitivity, or tactile defensive behavior, is contained in clinical reports of children with developmental delay, many of whom were born preterm. It also is seen in infants and children who otherwise appear normal. The behavior frequently is said to be a manifestation of sensory integration deficit and thought to have its origins in the prenatal or perinatal period. It appears as an infant’s overreaction to touch, generally the hands or oral-facial regions. With oral hypersensitivity, the infant may withdraw, gag, or retch when touched, even around the outside of the mouth. Some infants are intolerant of food with texture and resist transition from liquids or very smooth puree. Infants also may be hypersensitive to touch on their extremities, with prolonged palmar-mental reflex, exaggerated
hand and toe grasp, or leg withdrawal. An extreme case was a 2-month-old (corrected age) infant who, when supine, arched his buttocks off the table surface in response to his legs being grasped. Additional manifestations of tactile sensitivity may appear as an intolerance for grasping toys or handling play materials of certain textures. In another extreme example, a 1-year-old infant would gag when his hand was placed in dry macaroni. Some children are intolerant of normal clothing and even may avoid body contact. Such aversion adversely affects parent-infant bonding. The link between early tactile disturbances and learning disabilities at school age is unlikely to be a causal one, but may be related to a similar mechanism of brain dysfunction.
hand and toe grasp, or leg withdrawal. An extreme case was a 2-month-old (corrected age) infant who, when supine, arched his buttocks off the table surface in response to his legs being grasped. Additional manifestations of tactile sensitivity may appear as an intolerance for grasping toys or handling play materials of certain textures. In another extreme example, a 1-year-old infant would gag when his hand was placed in dry macaroni. Some children are intolerant of normal clothing and even may avoid body contact. Such aversion adversely affects parent-infant bonding. The link between early tactile disturbances and learning disabilities at school age is unlikely to be a causal one, but may be related to a similar mechanism of brain dysfunction.
Intrauterine Experience
The fetus is housed in a thermoneutral, fluid-filled space that is a source of cutaneous input throughout the body surface. Fetal movement provides tactile self-stimulation. Perhaps even more important, fetal movement often evokes a contingent maternal response. As term approaches and the intrauterine space becomes more constraining, the normal posture of flexion evokes hand-to-mouth, skin-to-skin, and body-on-body tactile feedback. The effect is progressive throughout gestation.
After a normal term birth, a ventral-to-ventral position is preferred by both mother and infant, with touch followed by slow stroking (37). Traditionally, the infant is then swaddled and held. As before birth, human proximity produces contingent touch.
Touch And Handling in the Neonatal Intensive Care Unit
After premature birth, tactile input is radically altered. The extrauterine fetus is frequently nursed naked, with the exception of maybe a diaper and a hat. The surface of the mattress generally is unyielding. She or he is exposed to air currents, cold stress, tape, instruments, handling by caregivers, and painful stimuli. Pressure is not uniform.
The type and frequency of tactile stimulation imposed on a preterm newborn in the NICU would be overwhelming even for a healthy adult. During a 2-week period, a sick neonate may be handled by more than 10 different nurses, in addition to physicians, occupational or physical therapists, laboratory and x-ray technicians, and, finally, the parents (38).
Handling occurs more often among the sickest infants, typically is related to procedures, generally is disturbing, and often is painful. In spite of wide attention in the literature to the consequences of excess handling in the NICU, the amount had not decreased from 1976 to 1990, even among the more severely ill infants (39). On the average, sick preterm newborns are handled more than 150 times per day, with less than 10 minutes of consecutive uninterrupted rest (26). Disturbance of sleep has biologic and immunologic consequences (40,41). Secretion of cortisol and adrenaline normally is inhibited during sleep. Growth hormone, which is released during quiet sleep, increases protein synthesis and mobilization of free fatty acids for energy use. Thus, sleep facilitates healing.
Excess handling has other significant physiologic consequences for the sick neonate. Blood pressure changes, alterations in cerebral blood flow, and episodes of oxygen desaturation are associated with noxious procedures, handling, or crying (42,43,44,45,46). Fluctuations in blood pressure may contribute to intracranial hemorrhage in the unstable preterm infant (47). Importantly, when caretakers monitor the infant’s level of oxygenation during procedures, the severity of hypoxemic episodes can be reduced significantly (42).
In addition to the hemodynamic impact of obviously noxious procedures described earlier, more benign manipulations, such as those that occur during neurodevelopmental assessment, also may adversely affect the preterm infant (25,46). Decreased plasma growth hormone has been reported after administration of the Brazelton Neonatal Behavioral Assessment Scale to preterm infants at 36 weeks PCA (48). Even at the time of discharge, the evaluation was associated with elevated cortisol levels (17,49). It is not clear whether these effects were from the neurodevelopmental assessment or from the stress associated with crying, which normally occurs during administration of the Neonatal Behavioral Assessment Scale. Thus, handling could be stressful even for stable preterm infants.
Tactile Intervention in the Neonatal Intensive Care Unit
The two general approaches to tactile intervention in the NICU provide either reduction of general handling or provision of planned touch experiences. Touch may be pressure alone or may include stroking. The neonates may be acutely ill or medically stable. These distinctions are important.
As part of an individualized approach to developmental care for acutely ill preterm infants, Als and colleagues (50) provided “minimal handling” and clustering of routine procedures, as well as positive tactile input and containment from bunting, rolls, positioning, and the like. Short- and long-term outcomes were improved for her intervention group compared to a nonintervention group. This technique has had a major impact on developmental intervention as a whole. Although it is not possible to isolate which aspect of this multimodal approach was effective, a significant change was made in the role of the nurse/caregiver.
More specific to the tactile modality, Jay (51) evaluated the effects of planned, gentle, touch for 12-minute periods four times a day for acutely ill preterm infants. This intervention, which consisted of hands-on contact but not stroking or manipulating, was associated with a lower fraction of inspired oxygen (FIO2) after 5 days compared to a similar nonintervention group.
In apparent contrast, Field and colleagues (52) and Scafidi and colleagues (53) initiated a touch intervention that differed from that of Jay (51) and Als and colleagues (50) in several important respects. The infants were recruited when “stable and growing” rather than during an acute period.
The treatment provided three 15-minute periods a day of massage (i.e., stroking and passive limb movement) for a 10-day treatment period. The massaged infants showed a greater weight gain, even though the groups did not differ in formula intake. These early effects on growth are thought to be mediated by induction of catecholamine release (48). The massaged infants also spent more time awake, showed better performance on the Brazelton neonatal assessment, were discharged to home 6 days sooner, and had better performance on developmental assessment at 8 months past term.
Although overall effects of the tactile intervention have been positive, the response of individual infants is variable and deserves attention. For example, an increase in periods of oxygen desaturation during parent touch, compared to baseline, has been reported (54). A number of infants have responded with apnea and bradycardia to the type of intervention described by Field and colleagues (52) and Scafidi and colleagues (53). This physiologic response may occur after the intervention. It is not clear whether the intervention itself was excessive or whether the abrupt shift after the cessation of tactile input led to an unexpected physiologic response. Individual differences exist among caregivers that also affect the infants’ responses to the stimulation. Finally, the distinction between touch, stroking, and massage is probably important.
All of these issues advise against standardized protocols of stroking or massage for all preterm infants. When any form of tactile and kinesthetic intervention is applied, the caregiver should continue to monitor the infant’s physiologic and behavioral responses before onset, during, and after the cessation of the intervention. Consistency of contact across caregivers would help.
Soft swaddling or clothing may provide tactile input in a more sustained fashion than periodic hands-on contact. Arguments that this obscures the view of the infant and interferes with temperature regulation by servocontrol simply argue against servocontrol rather than against covers. Clothing or swaddling actually could dampen the fluctuations in temperature that occur during incubator care and inhibit exaggerated movement or agitation, which in itself may be stabilizing. Creative swaddling with crisscrossed strips and Velcro could allow for more visualization of the infant. The use of swaddling in the NICU is necessary if the parent is to retain that option at home as a means of calming and supporting sleep. Swaddling is difficult to initiate after a period of not being swaddled.
Thus, the amount and type of stimulation are important and change with acuity, maturation, and the response of the infant. Parents need specific guidance and modeling from the beginning. The general order of tactile intervention might be:
If acutely ill—minimal handling, containment (e.g., swaddling, rolls), and gentle touch (e.g., warm hand) without stroking
When medically stable—holding, rocking gently, stroking, contining to swaddle
Minimal handling protocols have a definite place in the NICU, but not as the end point. It also is important during the hospital stay to help the infant develop increased tolerance for social contact and gentle handling, especially as discharge approaches. Systematic desensitization may even be necessary in cases of chronically ill infants.
Nonnutritive Sucking
Nonnutritive sucking is an important oral-tactile intervention that supports both feeding and early behavioral regulation. It represents an early endogenous rhythm and a manifestation of sensorimotor integration (55). As such, it is reported in the fetus (56) and observed in the preterm newborn before 28 weeks of gestation. The number of sucks per burst increases with maturation, whereas the duration of burst is fairly stable across ages.
Nonnutritive sucking experience may facilitate important physiologic and behavioral mechanisms and potentially reduce cost of care. Infants provided with nonnutritive sucking during gavage feeding showed significantly improved gastrointestinal transit time, greater suck pressure, more sucks per burst, and fewer sporadic sucks. They initiated bottle-feeding earlier, showed better weight gain, and thereby had shorter hospital stays (57,58). Having a pacifier continuously available, however, may not be beneficial and may, in fact, encourage inappropriate sucking patterns, particularly in the chronically ill neonate.
Nonnutritive sucking also acts as a behavioral organizer or facilitator. It has been shown to decrease motor activity and increase quiet states in stable preterm infants (59). It dampens an infant’s behavioral response after a painful procedure such as circumcision or heelstick (60,61), although it does not appear to dampen the cortisol response (17). It is noteworthy that sucking on a pacifier before the onset of repeated painful procedures, such as heelsticks, may be inappropriate, because aversive conditioning to the pacifier could occur.
Nipples used for nonnutritive sucking abound, varying in size, configuration, consistency, and utility. A feeding nipple is not designed for nonnutritive sucking and is inappropriate. It readily collapses; the infant experiences little resistance to his or her suck and may loll the device. Gauze inserted in the nipple may absorb oral secretions and breed bacteria. In larger infants, the nipple is unsafe because of possible aspiration. A variety of commercially available pacifiers should be available in any NICU to suit the individual needs of each infant. Some neonates who are hypersensitive to touch in the perioral region often respond positively to contact (and perhaps smell) from their own hands.
Vestibular System
The vestibular system, situated in the nonauditory labyrinth of the inner ear, responds to movement as well as directional changes in gravity. The three fluid-filled semicircular canals, one for each major plane of the body, lie at right angles to each other. The ampulla, located at the end of each canal, contains hair fibers in a sac, or cupula. Motion of the body or head causes pressure changes that move the cupula, which stimulates the hair cells and transmits an impulse along the vestibular portion of the eighth cranial nerve to the vestibular nuclei of the medulla. The vestibular organs consist of the utricle and saccule, which respond to changes of head position involving linear motion. The macula, a thickening in the wall of the utricle and saccule, contains hair cells sensitive to the position of the head. Impulses from the macula transmit along the vestibular nerve to the medulla and cerebellum. From there, information is transmitted to motor fibers going to the neck, eye, trunk, and limb muscles. There are no connections to the cortex (63). Vestibular stimulation affects level of alertness. Slow, rhythmic, continuous movement induces sleep. Periodic or higher amplitude swing increases arousal.
Development
Initial vestibular development is concurrent with auditory development, emanating from the same otocyst early in gestation. The three semicircular canals begin to form before 8 weeks of gestation, reaching morphologic maturity by 14 weeks, and full size by week 20 (29). The vestibular sacs probably develop at the same time. Response to vestibular stimulation has been observed by 25 weeks of gestation (33). The traditional vertex presentation of the fetus at term gestation is thought to occur from fetal activity in response to vestibular input.
Disturbances
Considerable research with animals has demonstrated the importance of both tactile and vestibular input (35,63). Lack of normal vestibular stimulation in the developing organism is thought to affect general neurobehavioral organization (31). Children who were born preterm are reported to have deficits in balance at preschool age (8), but this is not necessarily a vestibular problem.
Intrauterine Experience
The fetus experiences both contingent and noncontingent vestibular stimulation that varies during gestation. From the beginning of embryonic life, the fluid environment of the womb provides periodic oscillations and movements that emanate from normal movements of the mother as well as activity of the fetus itself. Reports by mothers of fetal movement occur around 16 weeks. After 28 weeks of gestation, there is a decrease in the relative amount of amniotic fluid, and, thus, the movement of the fetus becomes partially constrained by the more limited physical space. Vestibular experience is then less contingent on self-activation and more related to normal maternal activity and position change, which often occurs in response to fetal activity. In general, maternal activity level slows as parturition approaches.
After birth, the infant is held normally. Movement is slow from maternal breathing and shifting. Change of position is gradual, even by experienced parents. Vestibular stimulation is used to affect state—moving to upright or laying down increases arousal; monotonous side-to-side rocking and walking in the form of parental pacing reduce the level of arousal.
Vestibular Experience in the Neonatal Intensive Care Unit
Vestibular stimulation after preterm birth is limited to efficient manipulation or turning of the neonate by the caregiver. It clearly lacks any of the temporal qualities or contingencies that the maternal environment may have provided. Spontaneous limb movement generally is diffuse, often unrestricted, and typically disorganizing in its effect.
Intervention in the Neonatal Intensive Care Unit
Like the tactile sense, the early development of the vestibular system provides a theoretical basis for primary intervention with preterm neonates. More than 3 decades ago, Neal (64) demonstrated that daily rocking facilitated the development of preterm infants. Subsequent research simulated the intrauterine environment and provided compensatory vestibular stimulation. An oscillating waterbed was devised, which moved with the rhythm of maternal respirations but with an amplitude of less than 2.5 mm at the surface of the unoccupied waterbed. The safety of this paradigm, as well as efficacy in reduction in apnea of prematurity, has been well demonstrated (65). In addition, the infants on waterbeds demonstrated more organized sleep state and motor behavior, decreased irritability, enhanced visual alertness, and improved somatic growth (65,66,67,68).
Difficulty in nursing sick infants on an oscillating surface may have precluded more widespread adoption of the waterbed. An infant’s own movement also can induce more than optimal movement of the surface, as, for example, an infant with gastroesophageal reflux; however, some babies still are likely to benefit. At the least, it would seem prudent to provide a trial on a waterbed for a nonventilated infant with apnea of prematurity, before introduction of pharmacologic intervention.
Other sources of vestibular stimulation, such as rocking chairs, swings, and hammocks, have not been investigated formally. Rocking chairs probably belong in any nursery. Swings are questionable, given the excessive upright position of the baby and the standard rate of oscillation (i.e., too fast). A crib has been devised that provides controlled
motion similar to a woman walking; the duration of motion may be individually controlled and proportionally reduced over time (69), but the rate of oscillation appears too rapid for a preterm infant. The device appears to be effective in modulating fussiness in full-term infants and is used with preterm infants. The infants are well swaddled and further contained on each side by rolls.
motion similar to a woman walking; the duration of motion may be individually controlled and proportionally reduced over time (69), but the rate of oscillation appears too rapid for a preterm infant. The device appears to be effective in modulating fussiness in full-term infants and is used with preterm infants. The infants are well swaddled and further contained on each side by rolls.
Positioning
The physical position of an infant is part of the NICU tactile-vestibular experience. Nursing sick preterm infants routinely has been with the infant in the supine position and exposed, which may simplify management but may not be advantageous for the infant. Prone positioning in the NICU has been strongly supported physiologically. The current NICU dilemma is that the prone sleep position is contrary to the recommendation by the American Academy of Pediatrics (AAP), which now supports supine positioning because epidemiologic data associate supine positioning with a lower rate of sudden infant death syndrome. The optimal position of the infant needs to address anatomic and physiologic consequences. Positioning for optimal care in the NICU needs to take account of the AAP recommendation before the infant is ready for discharge to home.
Yu (70) demonstrated that gastric emptying was facilitated in either the prone or right lateral position compared to the supine or left lateral position. This was particularly significant for the sick preterm who already showed a delay in gastric emptying. The prone position, compared to supine, is associated with more quiet sleep and less active sleep or crying. Quiet sleep, in turn, is associated with improved lung volume, more stable respiration, less apnea, and improved PAO2 (71,72). Finally, the prone position compared to supine is associated with a higher PAO2 among healthy preterm infants and, even more significantly, in those with respiratory distress syndrome (72,73). The evidence suggests that, when possible, the sick infant should be nursed in a prone or right lateral position.
Parents of preterm infants often complain that their baby’s feet turn out. In fact, the legs more often are externally rotated at the hip. Grenier (74) described hip deformities seen on x-ray of preterm infants after prolonged nursing in a frog-leg position. The bulk of the diaper in extremely preterm infants exacerbates the problem. Winging of the scapula also is frequent in the preterm infant. Proper support of the trunk and limbs in the prone or supine position lessens this extreme rotation and may diminish orthopedic or neuromuscular complications.
In the prone position, placing the infant on a small folded strip from shoulder to hip, could allow more physiologic flexion and adduction. In side lying, it may be easier to position the infant in soft flexion. Gentle containment of the limbs usually can be managed with strips of soft cloth across the upper arm and thigh. Some movement should be allowed within a controlled range. A posture of physiologic flexion and adduction in the supine position requires swaddling. Maintenance of postures can be facilitated by nesting the infant in soft rolls, but the rolls must not reach above the level of the shoulder. Rolls placed at the buttocks don’t allow for adequate leg/hip extension. Each posture should facilitate the infant bringing hands to mouth.
For older, more medically stable preterm infants, infant seats are used as an alternative to continuous lying in bed. The infant should be swaddled and nested, the angle probably no greater than 30 degrees, and the length of time should be limited. Oxygen desaturation has been reported in stable preterm infants placed in car seats.
Kangaroo Care
Kangaroo care is a technique that evolved primarily in South America (75). Traditionally, the infant is clad only in a diaper and placed under the mother’s clothing between her breasts, remaining there according to the mother’s comfort, and feeding on demand. The technique provides fairly sustained multimodal stimulation: tactile, vestibular, proprioceptive, olfactory, and auditory. It appears to be safe for larger preterm infants or those who are medically stable. Temperature regulation in the infant does not appear to be a problem, but needs to be carefully monitored on an individual basis. It seems to have the greatest benefit in terms of facilitating and maintaining lactation and enhancing maternal sense of competency for these infants. The studies, however, are of insufficient sample size to evaluate whether morbidity, such as intracranial hemorrhage, is increased. More data are needed among medically stable infants before kangaroo care should be attempted prior to 32 weeks conceptional age or with infants requiring mechanical ventilation. The increased tactile stimulation and additional handling easily could be overly stressful for the immature or sick infant.
Chemical Senses
The chemoreceptors include taste and olfaction. Taste receptors are in the taste buds, which are located primarily in the papillae of the tongue but also are found on the soft palate and epiglottis (29,76). Taste stimuli (i.e., sweet, sour, bitter, salt) transmit to the brainstem with a primary branch to the hypothalamus. Cortical regions are involved in learned taste preferences. The olfactory receptors are located in the lining of the olfactory epithelium in the posterior portion of the nasal passage. The afferent pathway has no cortical projection area, but is direct to the limbic system. Olfaction plays an important part in gustatory experiences. Olfaction also is an integral part of infant attachment to the caregiver and may even be mutual (77).
Development of Taste
The chemoreceptors are well developed within the first trimester (29,76). Taste buds appear around 8 to 9 weeks of gestation. The receptors are present at least by week 16
of gestation, and increase by term to adult levels. During the second half of gestation, morphologic changes occur that continue after term. Taste receptors are functional before birth. Injection of distinct tastes into the amniotic fluid of pregnant women between 34 and 39 weeks of gestation alter fetal swallowing behavior, which increases with the sweeter taste and decreases with bitter taste (29).
of gestation, and increase by term to adult levels. During the second half of gestation, morphologic changes occur that continue after term. Taste receptors are functional before birth. Injection of distinct tastes into the amniotic fluid of pregnant women between 34 and 39 weeks of gestation alter fetal swallowing behavior, which increases with the sweeter taste and decreases with bitter taste (29).
Taste discrimination has been measured by differential consumption, autonomic responses, and the presence of characteristic facial expressions. Plain water evokes an aversive response, which may be a biologically based protective mechanism. Taste is sufficiently sensitive at term to detect a 0.1 mol/L-concentration of NaCl in water (78). Full-term neonates, even anencephalic infants, demonstrate differential behavioral responses to sweet, bitter, sour, and salt (79). In a behavior described as “savoring,” normal newborns discriminate between different concentrations of sucrose and even among various sugars (78). Preterm infants (30 to 36 weeks of gestation) show stronger sucking in response to glucose, compared to plain water, and characteristic behavioral expressions in response to sour or bitter solutions (80). Behavioral response to formula, or breast milk, administered to the tip of the tongue has been documented in preterm infants prior to 28 weeks of gestation (Zorc L. unpublished doctoral dissertation, 2000).