Main

Any determination of mechanisms underlying the sudden infant death syndrome (SIDS), the sudden and unexplained death of an infant occurring in the 1st year of life, must consider some of the known characteristics associated with the syndrome. These characteristics include enhanced risk with the prone sleeping position (1, 2), a substantially increased incidence following prenatal or postnatal tobacco exposure (3), a temporal association with sleep, periods of tachycardia before the fatal event (4), and an incidence confined principally between the 2nd and 4th months of life. Less well-documented associations, but findings repeatedly encountered in reports, are profuse sweating (5) and abnormally high core or environmental temperatures (6). Among the characteristics associated with the fatal event, a remarkable, often short-lasting bradycardia, accompanied by hypotension, sometimes in the presence of continued respiratory efforts, has been reported (7, 8).

The findings of tachycardia and profuse sweating before death implicate initial exaggerated sympathetic nervous system activity, while bradycardia and hypotension during the fatal event suggest that a subsequent sympathoinhibition, but parasympathetic nervous system recruitment accompany the fatal event; this sequence parallels the trend of events that are associated with shock from blood loss or deep pain (9). The possibility of a perfusion failure represents a departure from current interpretations of the mechanism of death as resulting primarily from a respiratory collapse. The sudden bradycardia also argues against particular arrhythmia failures, such as prolonged Q-T intervals (10) that normally result in death by ventricular fibrillation. Although a proportion of SIDS deaths may result from suffocation, ventricular fibrillation, CO2 intoxication, or hypoxia, a proportion may also result from a sympathoinhibition, precipitated by a shock or shock-like scenario. The significant element in this scenario is a profound loss of blood pressure, and an apparent inability to restore or induce compensatory responses to assist restoration of vascular tone. It may be the case that the initial hypotension may be triggered by a number of mechanisms other than the classic shock sequence of sympathoexcitation followed by sympathoinhibition. Marked blood pressure falls occur spontaneously during rapid eye movement sleep, for example, and hypotension follows sustained hypoxia.

The range of circumstances surrounding the fatal event in SIDS suggest a heterogeneous group of disorders, some of which appear to be of a respiratory origin, eg those associated with infant faces embedded in restricted air space of pillows or bedding (11). The airflow restriction may be of an internal obstructive nature, as indicated by increased incidence of obstructive apnea in infants who succumb (5). Other fatal events appear to be of cardiovascular nature, eg bradycardia and hypotension during the fatal event (8); the different scenarios suggest that infants may succumb through several failure modes, and that the capability to adequately recover from a challenge may be a more significant issue for survival than the particular mechanism of failure. Arousal mechanisms from sleep have been implicated as a common vehicle for recovery from compromising events (12), and are undoubtedly a significant issue in restoration of compensatory forebrain and other influences that assist recovery of vital processes. The concept of “arousal” should perhaps be extended to include restoration or activation of somatic muscle tone in addition to transition of sleep state. Recruitment of muscle tone is essential for respiratory challenges, eg head turning from an obstruction, enhanced tidal volumes to hypercapnia or hypoxia, or appropriate responses to a blood pressure loss. This last interaction implicates a role for somatic musculature on autonomic functioning (sympathetic control). The relationships derived from animal studies suggest that effective compensatory responses to loss of blood pressure involve somatomotor action, and especially the respiratory somatic musculature to restore blood pressure. Cats subjected to blood loss, for example, successfully respond to the hypotension and are able to restore blood pressure by enhanced inspiratory and expiratory efforts, tachypnea, and exaggerated extensor skeletal muscle action (13). Those successful compensatory motor responses are associated with substantial activity changes on the rostral ventral medullary surface, an area classically implicated in blood pressure regulation. However, the recovery efforts obviously also recruit motor areas that receive signals from central structures indicating low blood pressure, and activate muscles for recovery; the recovery response is particularly apparent for respiratory muscles (14).

Neural mechanisms responsible for compensatory responses to a hypercapnic, hypoxic, or bradycardic/sympathoinhibitory challenge may involve brain sites not normally considered as “respiratory,” “arousal,” or “cardiovascular” areas. A principal focus in respiratory control studies has been to determine brain areas responsible for central pattern generation underlying regular breathing rhythm, and that attention has usually been directed to the medulla. Issues of blood pressure control similarly focus on medullary reflex loops. However, a major interest for SIDS research is determining which brain structures are recruited to restore breathing when pattern generators fail, ie from an apneic episode, or which neural sites compensate for blood pressure falls incompatible with survival; those structures may lie outside normally considered medullary sites. Recent evidence from functional magnetic resonance studies in normal subjects (15) and in children afflicted with congenital central hypoventilation syndrome (16), as well as electrophysiological recording and lesion studies in animals, implicate portions of the cerebellum, especially the cerebellar fastigial nucleus, in modulating appropriate responses to O2 and blood pressure challenges. Bilateral lesions of the fastigial nucleus result in a fatal progression after blood pressure lowering (17, 18), or uncorrected influences from hypercapnia on breathing (19). Regions within the cerebellum are often associated with an “error correction” role, ie the correction of appropriate motor output after sensing of aberrant afferent signals. That “error correction” is classically associated with motor performance, and tested clinically with a motor task, such as finger pointing. Comparable regulatory roles for the cerebellum apparently extend to blood pressure and breathing control (15).

Cerebellar compensatory reactions to hypotension or hypertension are likely mediated through afferent activity from the inferior olivary nucleus via climbing fibers to Purkinje cells and the fastigial nucleus of the cerebellum, and output to vestibular sympathetic pathways, as well as somatomotor regions. Portions of the inferior olive, a major input relay to the cerebellum, show significant gliosis in infants who have succumbed to SIDS, and isolated cases of inferior olive hypoplasia are associated with profound respiratory dysfunction (20). More recent findings indicate deficiencies in SIDS victims in muscarinic and kainate receptors within the ventral medullary surface (21, 22) and serotonergic receptors in medullary sites including the inferior olive and caudal medullary raphe regions associated with hypotension and sympathoinhibition (23, 24). The inferior olive also shows increased c fos expression to a variety of manipulations that trigger vasodepression (Bandler and Keay, Department of Anatomy and Histology, University of Sydney, NSW, Australia, personal communication, 1999). Delayed maturation of cerebellar regions has been observed in SIDS victims (25). Vestibular contributions to blood pressure regulation are well-known, since body positioning requires rapidly acting compensation of regional blood pressure to assist appropriate perfusion, frequently operating in a “feed-forward” or anticipatory fashion (26, 27). Deficiencies in such regulation are frequently observed clinically, for example, in rapid rising from a supine to vertical position.

A cerebellar/vestibular role for compensatory responses to cardiovascular or breathing challenges may participate in the position-dependent risk factor for SIDS. Afferent position information from the position-sensitive receptors travel by way of the vestibular nuclei and inferior olive to the cerebellum, which then projects to reticular and rostral ventrolateral medullary sympathetic areas via vestibular nuclei (27). Such vestibular contributions may be the significant factor underlying the substantial prone versus supine position differences in cardiovascular and breathing control in low birth weight infants (28); the blood pressure response to head-up tilt is greatly reduced in full-term infants in the prone position as well (29). Cerebellar-mediated vestibular input from rocking also reduces obstructive apneic events (30).

A substantial number of cerebellar developmental abnormalities have been identified, including intracerebellar hemorrhage (31); some of these aberrations are associated with marked respiratory disturbances; other case reports describe significant cardiovascular disturbances following midline cerebellar stimulation in humans (32). If cerebellar abnormalities are present in SIDS victims, the resulting physiologic characteristics are subtle, since cardiac and respiratory patterns that differentiate future SIDS victims from controls can only be identified with careful attention to partitioning breathing and cardiac patterns by sleep state and time-of-night (33). However, the pronounced influences on vital functions from cerebellar syndromes emphasize the potential capacity for more-modest cerebellar deficiencies to disturb physiologic functions. Animal data suggest that cerebellar mechanisms are particularly recruited during extreme challenges, rather than during routine regulation of breathing and blood pressure (17, 19).

The epidemiologic evidence for SIDS suggests a prominent role for prenatal nicotine exposure and low maternal hematocrit (34). Recent animal evidence demonstrates the importance of adrenal catecholamine release on autoresuscitation from hypoxia, and that prenatal nicotine exposure compromises autoresuscitation, possibly by interference with cardiac conductance changes associated with altered adrenal catecholamine outflow (35–37). The potential role of loss of protective vagal influences on cardiac conduction disturbances has been outlined earlier (38).

However, prenatal or postnatal central damage could also affect compensatory recovery mechanisms. Cerebellar Purkinje cells receive afferents from the climbing fibers of the inferior olive and are especially sensitive to neurotoxic damage from agents such as harmaline (39), perhaps because of the unique limited geometry of the afferent system. It appears that the peculiar arrangement of Purkinje fibers poses a risk for excitotoxic injury from a variety of challenges, including hypoxic exposure, nicotine, or harmaline. Harmaline exposure results in aberrant responses to blood pressure challenges that are comparable to lesions of the fastigial nucleus of the cerebellum. In rats, it appears that such disturbances are time-dependent, with recovery by alternate pathways several weeks after damage (40). It may be the case that repetitive hypoxic insults occurring postnatally, or fetal damage to this blood pressure/breathing regulatory system from nicotine exposure establishes a less-than-optimal system for responding to ventilatory or blood pressure challenges; the time-dependent response capability may be an issue, because of the narrow window of SIDS risk (between the second and fourth months of life).