Breath-Holding Spells and Reflex Anoxic Seizures

Published on 13/04/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 8863 times

Chapter 64 Breath-Holding Spells and Reflex Anoxic Seizures

Breath-holding spells and reflex anoxic seizures are nonepileptic paroxysmal events. The events are benign, but can be frightening to parents and others observing an episode. It is important to differentiate these episodes from epileptic seizures so that the child is not inappropriately treated with antiepileptic medication.

Possibly the earliest report of breath-holding spells was published in 1737 by Nicholas Culpepper, who gave the following description: “There is a disease … in children from anger or grief, when the spirits are much stirred and run from the heart to the diaphragms forceably, and hinder or stop the breath … but when the passion ceaseth, this symptom ceaseth.”

The clinical characteristics were well recognized and described in the pediatric literature in the 19th and early 20th centuries. More recent reports have provided a better understanding of the pathophysiology of these events.

Breath-Holding Spells

Clinical Features

The term breath holding is a misnomer and implies that the child is voluntarily holding his or her breath in a prolonged inspiration. Breath-holding episodes actually occur during expiration and are involuntary. Breath-holding spells are not uncommon, with an incidence of 4.6–4.7 percent [Linder, 1968; Lombroso and Lerman, 1967]. The typical age of onset is between 6 and 18 months, although occasionally the onset may occur in the first few weeks of life [Breukels et al., 2002]. Fewer than 10 percent have onset after 2 years of age [DiMario, 1992]. The frequency of episodes ranges from several times daily to once yearly. The spells are often spaced weeks to months apart at onset, and increase in frequency to as many as several per day during the second year of life [Laxdal et al., 1969; DiMario, 2001]. Breath-holding spells are classified by the color change manifested in the child during an event. Cyanotic episodes are more common than pallid episodes. In some instances, there are features of both cyanosis and pallor, and these are termed mixed episodes.

Cyanotic breath-holding spells are often precipitated by emotional stimuli, such as anger or frustration. The child typically cries vigorously but usually for less than 15 seconds, then becomes silent, and holds the breath in expiration. The apnea is associated with the rapid onset of cyanosis. Some episodes may resolve at this point, but there may be loss of consciousness and a brief period of limpness, followed by opisthotonic posturing. Recovery is usually within 1 minute, with the child having a few gasping respirations and then a return to regular breathing and consciousness.

Pallid breath-holding spells are usually provoked by sudden fright or pain. A fall with a minor injury to the head is frequently the precipitating event. An unexpected event or a surprise seems to play a role in triggering the spell. Sometimes the provoking event is not witnessed, and the child is found already in an episode. The child may gasp and cry, although it is usually for only a brief period of time. The child then becomes quiet, loses consciousness, and becomes pale. Limpness and diaphoresis are common. Clonic movements of the extremities and incontinence may occur with more severe episodes. Cyanosis may occur during the episode but is much milder than with cyanotic breath-holding spells. The child typically regains consciousness in less than 1 minute but may sleep for several hours after the episode.

An association between behavior problems, emotional factors, and breath-holding spells has been discussed by many investigators. Breath-holding spells were described by Abt [1918] as occurring in “neuropathic children of neuropathic parents.” Bridge et al. [1943] stated that children susceptible to breath holding are usually of the active, energetic type, who react vigorously to situations, and that episodes were precipitated by “spoiled child reactions.” Breath-holding spells were felt to be a sign of a disturbed parent–child relationship by Kanner [1935]. Laxdal et al. [1969] reported that 30 percent of the children with breath-holding spells had abnormal behavior, including temper tantrums, hyperactivity, and stubbornness. To investigate the role of behavior and breath holding further, DiMario and Burleson [1993b] studied behavior in children with breath-holding spells compared with controls and found no differences in the behavioral profiles, suggesting that breath-holding spells are nonvolitional and cannot be equated with a temperamentally difficult child.

Breath-holding spells generally decrease in frequency during the second year of life. By 4 years of age, 50 percent of children will no longer have episodes. Almost all will have stopped having episodes by age 7–8 years [DiMario, 1992; Goraya and Virdi, 2001]. Syncopal episodes occur in late childhood or adolescence in as many as 17 percent of patients with breath-holding spells [Lombroso and Lerman, 1967].

Serious complications with breath-holding spells are rare. Taiwo and Hamilton [1993] reported a prolonged cardiac arrest in a patient with breath-holding spells. The few reported deaths may have been precipitated by aspiration or occurred in children who were at the severe end of the spectrum of breath-holders, often with structural abnormalities of the respiratory tract or complicated medical histories [Paulson, 1963; Southall et al., 1987, 1990].

Clinical Laboratory Tests

A detailed history of the event, including the precipitating circumstances, is essential in making the diagnosis of breath-holding spells. If the event was not witnessed from onset, important details may not be available. A video recording by the parents may be helpful in confirming the diagnosis. Usually, no laboratory tests are needed to make the diagnosis. An electroencephalogram (EEG) is usually not indicated, unless the convulsive activity is prolonged or the clinical description is incomplete and epileptic seizures cannot be ruled out. If ocular compression is performed in patients with pallid breath-holding spells, there may be asystole on cardiac monitoring, and slowing or suppression of voltage on EEG [Lombroso and Lerman, 1967; Stephenson, 1978]. Long QT syndrome is rare but should be considered as part of the differential diagnosis in a child with breath-holding spells. Patients with long QT syndrome have episodes of loss of consciousness that may be induced by injury, fright, or excitement. An electrocardiogram should be considered in any patient with breath-holding spells [Breningstall, 1996; Franklin and Hickey, 1995].

Pathophysiology

Cyanotic Spells

The pathophysiology of cyanotic breath-holding spells is complex and not completely understood. Cyanosis occurs early in the episode, which is unusual during voluntary breath holding. In breath-holding spells, the breath is held in full expiration, which also is not typical with voluntary breath holding [Livingston, 1970]. Gauk et al. [1963] studied a child during a cyanotic breath-holding episode with cinefluorography and noted the diaphragm to be high, as would be seen in full expiration, and motionless during the period of apnea. Spasm of the glottis and respiratory muscles, with increased intrathoracic pressure, occurs during expiration. Increased intrathoracic pressure reduces cardiac output, causing a decrease in cerebral perfusion. Lombroso and Lerman [1967] suggested that violent crying could lead to hypocapnia, which would also impair cerebral circulation.

Southall et al. [1985] further evaluated the prolonged expiratory mechanism in nine infants with cyanotic episodes that were usually triggered by noxious stimuli. Arterial oxygen saturation fell below 20 mm Hg within 20 seconds. Loss of consciousness occurred after 30 seconds. Measurements of respiratory movements, airflow, and esophageal pressure, and, in some patients, microlaryngoscopy and chest fluoroscopy were obtained. They documented no inspiratory flow during the period of apnea but continued expiratory muscle activity at low lung volumes with partial or complete glottic closure. No intracardiac shunt could be demonstrated. The rapid fall in arterial oxygen saturation was attributed to lack of ventilation at a maximum expiratory position in the presence of a rapid circulation time. The researchers hypothesized that central and peripheral neural respiratory control was functioning normally but was interfered with by a mechanical defect involving lung-volume maintenance. This defect could occur because of an excessively compliant rib cage, allowing alveolar collapse. This collapse, in turn, could lead to stretching of the airways and their stretch receptors, inappropriately simulating maximum lung volumes and thereby inhibiting inspiration. Southall et al. [1990] did further evaluations of prolonged expiratory apnea with krypton infusion scans and demonstrated krypton outside the lung fields, without evidence of an intracardiac shunt. They felt there was intrapulmonary shunting that contributed to the rapid onset and severity of the hypoxemia.

The relation between breath holding and chemosensitivity has also been investigated. Anas et al. [1985] hypothesized that persons with cyanotic breath-holding episodes have blunted ventilatory chemosensitivity. Because of the difficulty of measuring chemosensitivity in toddlers, they measured ventilatory responses to progressive hypercapnia and to progressive hypoxia in subjects aged 11–50 who had a history of cyanotic breath-holding spells and compared the results with a control group. Contrary to their hypothesis, the majority of persons with a history of cyanotic breath-holding spells had normal ventilatory responses. However, no one with a history of breath-holding spells had high normal responses to hypercapnia or hypoxia, as did some individuals in the control group. They postulated that the difference between the groups might represent the vestige of a disorder of ventilatory chemosensitivity that resolved with maturation.

Kahn et al. [1990] also investigated the relation between breath holding and cardiorespiratory control. The study included 71 infants with a history of breath-holding spells and age- and gender-matched controls. The median age of infants in the study was 14 weeks, which is younger than the typical age for onset of breath-holding episodes. The infants with breath-holding spells were significantly more often covered with sweat during sleep and wakefulness compared with control infants. One-night sleep studies were obtained in each infant. The infants with breath-holding spells had significantly less non-rapid eye movement (REM) stage III sleep, more indeterminate sleep, more arousals, and more sleep-stage changes than the control infants. Airway obstructions during sleep occurred in 41 infants with a history of breath holding, compared with 6 in the control group. The obstructions were generally short and not accompanied by significant bradycardia or oxygen desaturation. The researchers concluded that there was a common underlying mechanism resulting in airway obstruction during breath-holding spells and sleep, which possibly involved the autonomic nervous system because the autonomic nervous system controls the patency of the upper airways. Guilleminault et al. [2007] performed polysomnography in 14 children with cyanotic breath-holding spells and found an abnormal respiratory index in all 14. Examination showed upper airway narrowing, and adenotonsillectomy was performed in 13 with marked improvement in sleep-disordered breathing and resolution of their breath-holding spells.

Kohyama et al. [2000] did polysomnography to evaluate REM sleep in seven children with breath-holding spells and nine normal age-matched controls. The children with breath-holding spells had a significant decrease in ocular activity during REM sleep, especially during the last third of the night, compared with the controls. Relative elevation of cholinergic tone, compared with monoaminergic tone, is considered to be involved in the physiologic increase of REM sleep in the later cycles of the night. The vestibular nucleus and the medioventral caudal pons are believed to be involved in bursts of eye movements during REM sleep. They hypothesized that there was a functional disturbance in the pons of children with breath-holding spells. The study also suggests that the autonomic nervous system is involved because of the more pronounced decrease in eye movement in the later cycles of the night, which are regulated by the autonomic nervous system.

DiMario and Burleson [1993a] used noninvasive methods to evaluate autonomic nervous system function in children with severe cyanotic breath-holding spells. Compared with controls, the breath-holders had a significantly greater increase in pulse rate at 15 seconds of standing after rising from the supine position. Breath-holders also had a greater decrease in diastolic blood pressure without an increase in systolic blood pressure after standing from the supine position. These results suggest that there is autonomic dysregulation in children with cyanotic breath-holding spells. Using the results of this study and prior pieces of work, DiMario and Burleson postulated that, in addition to evidence of parasympathetic excess, children with cyanotic breath holding exhibit subtle sympathetic excess, which mediates vascular resistance, arterial distensibility, and blood flow through the lungs. This sympathetic overactivity could cause the intrapulmonary shunting and subsequent hypoxemia [Southall et al., 1990].

Pallid Spells

Excessive vagal tone leading to cerebral hypoperfusion is the underlying cause of pallid breath-holding spells. Observation of children during a typical episode reveals marked bradycardia or asystole [Bridge et al., 1943]. Ocular compression that triggers the oculocardiac reflex has been used to evaluate vagal tone in children with breath holding [Lombroso and Lerman, 1967; Stephenson, 1978]. This maneuver results in transmission of afferent signals to the brainstem via the ophthalmic division of the trigeminal nerve and efferent parasympathetic signals via the vagus nerve. In 61–78 percent of children with pallid breath-holding spells, ocular compression resulted in asystole of 2 seconds or longer, compared with 23–26 percent of children with cyanotic breath-holding spells [Lombroso and Lerman, 1967; Stephenson, 1978]. Episodes that occurred spontaneously during cardiac monitoring were also associated with asystole [Lombroso and Lerman, 1967; Maulsby and Kellaway, 1964]. The asystole during spontaneous episodes is believed to be vagally mediated. When asystole is prolonged, a reflex anoxic seizure may occur.

The role of underlying autonomic dysfunction in children with pallid breath-holding spells has been investigated in a small number of patients. Measurements of mean arterial pressures, pulse rates, electrocardiograms, and plasma norepinephrine levels were obtained in patients and controls during changes in position. The breath-holders had a statistically significant decrease in mean arterial pressure and an unsustained increase in pulse rate during the prone to standing maneuver. One child with pallid breath-holding spells had a plasma norepinephrine level that was 60 percent below the mean for both groups [DiMario et al., 1990]. Further evaluation of autonomic function was performed in children with either pallid or cyanotic breath-holding spells. Respiratory sinus arrhythmia, which is an established measure of vagal tone, was measured. There were no significant differences between controls and children with cyanotic breath-holding spells. The children with pallid spells, however, had a marked difference in respiratory sinus dysrhythmia, with less variability compared with controls and those with cyanotic episodes [DiMario et al., 1998]. These studies suggest that there may be an underlying parasympathetic dysregulation in children with pallid breath-holding spells.

The role of anemia in the pathophysiology of breath-holding spells was suggested by Holowach and Thurston [1963]. They found that 23.5 percent of 102 children with breath holding had a hemoglobin level less than 8 g/100 mL, compared with 7 percent and 2.6 percent in two control groups. Some studies did not find any significant difference in hemoglobin levels in the breath-holding group, compared with the control group [Laxdal et al., 1969; Maulsby and Kellaway, 1964]. Kolkiran et al. [2005] reported that asystole during breath-holding spells was prolonged in children with iron deficiency. There are reports of children with breath-holding spells and concomitant anemia, who had resolution of their spells with correction of the anemia [Bhatia et al., 1990; Colina and Abelson, 1995; DiMario, 1992; Mocan et al., 1999; Orii et al., 2002]. Tam and Rash [1997] described a child with pallid breath-holding spells associated with transient erythroblastopenia of childhood. The spells resolved after treatment with iron but before the anemia resolved. Daoud et al. [1997] studied 67 children with breath-holding spells to investigate the effect of iron therapy. Treatment and placebo groups were similar with respect to gender, age at onset, and frequency and type of spells, and had similar blood indices, including packed cell volume, mean corpuscular volume, saturation index, total iron binding capacity, and serum iron. At the end of the treatment period, 51.5 percent of the children treated with ferrous sulfate had complete remission of spells, and an additional 36.4 percent experienced a greater than 50 percent reduction. No children in the placebo group had total remission of spells, and only 5.9 percent had a greater than 50 percent reduction. As expected, the treatment group experienced significant improvement in the hemoglobin level and total iron-binding capacity. However, some children who were not iron-deficient had a favorable response to iron therapy, and some who were iron-deficient did not respond. A recent study suggests that checking serum soluble transferrin receptor levels in children with breath-holding spells may be helpful in assessing iron status. An increase in serum soluble transferrin receptor levels is an early change seen in iron deficiency before anemia develops [Handan et al., 2005]. Iron deficiency may play a role in the pathophysiology of breath-holding spells because iron is important for catecholamine metabolism and neurotransmitter function [Daoud et al., 1997].

Genetics

In children with breath-holding spells, there is a positive family history of similar episodes in 23–38 percent, suggesting a genetic influence [Laxdal et al., 1969; Lombroso and Lerman, 1967]. An evaluation of family pedigrees found that 27 percent of 114 proband parents and 21 percent of proband siblings had a history of breath holding. Several families had some members with pallid spells and other members with cyanotic spells. The male to female ratio was 1:1.2 and the risk of transmission from parent to child was 50:50. There were seven instances of father to son transmission, ruling out an X-linked inheritance. Using a regression model for pedigree analysis, the inheritance pattern was consistent with an autosomal-dominant pattern with reduced penetrance [DiMario and Sarfarazi, 1997].

Treatment

The most important aspect of treatment of breath-holding spells is to reassure the family of the benign nature of the spells. It is important to emphasize that the episodes do not lead to mental retardation or epilepsy. Although parents are inclined to pick up a child who is having a breath-holding spell, they should be instructed to place the child in a lateral recumbent position so as not to prolong the period of cerebral anoxia. Initiation of cardiopulmonary resuscitation should be avoided. Although anger and frustration are often precipitants for breath-holding spells, parents should be encouraged not to alter customary discipline for fear of triggering an episode [DiMario, 1992]. Parenting a child with breath-holding spells has been associated with more maternal stress than parenting a child with a convulsive seizure disorder, and parents of children with breath-holding spells are at risk for developing dysfunctional parenting behaviors [Mattie-Luksic et al., 2000]. Referral of parents to professionals to help with stress and parenting skills should be considered.

Treatment with iron therapy should be initiated in any child who has iron deficiency anemia and should be considered in any child with breath-holding spells because children without anemia may have improvement in their breath-holding spells. The convulsive movements seen during breath-holding spells are reflex anoxic seizures, which are not epileptic and do not require antiepileptic treatment. There have been a few patients who have been reported to have prolonged seizures and even status epilepticus from breath-holding spells [Emery, 1990; Kuhle et al., 2000; Moorjani et al., 1995; Nirale and Bharucha, 1991]. It is presumed that these patients have a lowered seizure threshold and that hypoxia-ischemia triggered the seizures [Emery, 1990]. Stephenson [1990] has termed these events anoxic-epileptic seizures. Treatment with antiepileptic medication may stop the seizure activity but not the breath-holding spells. Atropine (0.01 mg/kg two or three times daily) is effective for pallid breath-holding spells, but its use is rarely warranted [McWilliam and Stephenson, 1984; Stephenson, 1980]. Piracetam, which has a chemical structure similar to gamma-aminobutyric acid (GABA), has been used to treat children with breath-holding spells. In a study of 76 children with breath-holding spells, treatment with piracetam for 2 months resulted in 92 percent having no recurrence of episodes for 6 months after treatment, compared with 30 percent who received placebo [Donma, 1998]. Azam et al. [2008] treated 52 children with breath-holding spells using piracetam in doses ranging from 50 to 100 mg/kg/day and iron supplementation in those with hemoglobin less than 10 g/100 mL. In 81 percent of the children the spells completely resolved, and in an additional 9 percent the frequency and intensity were reduced. Piracetam has not received approval from the U.S. Food and Drug Administration and is designated only as an orphan drug for use in myoclonus.

Reflex Anoxic Seizures

Reflex anoxic seizures are nonepileptic events resulting from cardiac asystole of vagal origin [Stephenson, 1990, 2001]. Pain and surprise are common provoking factors for the events [Stephenson, 1980]. Reflex anoxic seizures may occur with pallid breath-holding spells but also have been reported with minor blows to the occiput, expelling hard stools past an anal fissure, venipuncture, intramuscular injections, and seeing an intravenous scalp drip [Braham et al., 1981; Gordon, 1982; Lombroso and Lerman, 1967; Roddy et al., 1983; Stephenson, 1980]. Nonepileptic anoxic seizures may also occur after syncope, cyanotic breath-holding spells, or any event that results in a sudden reduction in cerebral perfusion or hypoxia.

Clinical Features

Reflex anoxic seizures occur a few seconds after the provocation and are characterized by loss of muscle tone initially and later by tonic posturing. There may be opisthotonic posturing in some patients. A few jerks at the onset and end of an anoxic seizure may occur and probably represent myoclonic phenomena. A snoring type of inspiration or snort occurring close to the restoration of the cardiac rhythm is often noted. Urinary incontinence happens in approximately 10 percent of children with anoxic seizures, with bowel incontinence occurring less commonly [Stephenson, 1990]. Other less common features include adversive head movements, limb quivering or twitching, agitation or fear, vomiting, and tongue biting. The color change seen with an anoxic seizure may be cyanosis or pallor, depending on the mechanism producing loss of consciousness. The duration of unconsciousness is almost always less than 1 minute. Most patients experience a rapid recovery of consciousness, but some will be dazed or disoriented for a short period. Some will be drowsy after an episode and may sleep [Stephenson, 1990]. Occasionally, patients will have prolonged seizure activity after syncopal spells. These events, termed anoxic-epileptic seizures, are epileptic seizures triggered by hypoxia in patients with a lowered seizure threshold [Stephenson, 1990]. A positive family history of epilepsy may make some children more prone to anoxic-epileptic seizures [Horrocks et al., 2005].

Pathophysiology

The mechanism of reflex anoxic seizures has been studied by using ocular compression with EEG and cardiac monitoring [Gastaut and Fischer-Williams, 1957; Gastaut and Gastaut, 1958; Lombroso and Lerman, 1967; Stephenson, 1978]. Ocular compression induced asystole in susceptible patients. If asystole lasted 3–6 seconds, there were no clinical symptoms and the EEG demonstrated only desynchronization. When asystole lasted 7–13 seconds, slow waves appeared, usually associated with altered consciousness. If asystole was prolonged for 14 seconds or more, there were often myoclonic jerks or tonic posturing. The EEG during this time reveals no electrocerebral activity. With return of cardiac activity, there was again high-voltage slow-wave activity on the EEG with return of normal activity over 20–30 seconds [Gastaut and Fischer-Williams, 1957]. At no time during EEG monitoring were epileptiform discharges present. Some patients have had spontaneous episodes or episodes triggered by other stimuli, such as venipuncture during EEG and cardiac monitoring, and have demonstrated similar changes to those seen with ocular compression [Braham et al., 1981; Gordon, 1982; Lombroso and Lerman, 1967; Roddy et al., 1983]. Ocular compression increases vagal tone, with the afferent pathway involving fibers from the trigeminal nerve originating from the cornea, iris, and eyelids. In contrast, episodes induced by exteroceptive stimulation, such as pain and emotion, have afferent fibers in various sensory pathways. In both situations, the vagal reflex centers are located in the brainstem in the nucleus ambiguus [Chen and Chai, 1976]. The efferent pathway involves the cardioinhibitory fibers of the vagus nerve.