These are universally present in normal people. They appear during the initial stages of sleep and also during rapid eye movement (REM) sleep. Typically, they manifest as slight contractions of face muscles and brief movements of the fingers or toes. There are no electroencephalographic (EEG) correlates.⁶⁹ Massive bilateral jerks, mainly involving the legs, can occur during the initial stages of sleep in association with a change in sleep state or arousal.¹¹⁴

This represents a rare physiologic phenomenon that usually occurs during and after muscular fatigue and when the linear or a particular group of muscles are placed in unsupported postures. The affected muscle groups may vary, but in any individual, each episode of myoclonus involves the same muscles. The myoclonic jerks last a few seconds to several minutes. There are no EEG changes accompanying the myoclonus, and a spinal origin has been suggested.⁶⁹ Even hiccoughs, considered a physiologic phenomenon, may occasionally be intense and simulate a paroxysmal disorder in small infants.

These have been called sursaut diurne in the French literature⁶⁹ and are sudden, bilateral myoclonic jerks that may be either tight or massive and appear as a surprise reaction to a sudden sensory stimulus. Although segmental myoclonus is rare in young children, startle responses are very common. The Moro reflex has a startle component, and some kind of startle response can be demonstrated in most normal infants. These normal physiologic responses must be distinguished from the exaggerated Moro reflex and pathologic stimulus-induced myoclonus, often with opisthotonus seen in children with static or progressive encephalopathy.⁵⁸

The majority of adults and children have nocturnal bruxism at some point in their lives.¹³ In children with autistic behavior and mental retardation, bruxism also occurs in the waking state and may lead to severe dental attrition.¹⁰⁹

More common disorders are considered in Chapter 276 (see also Hanson and Peck⁷⁹).

In 1982, Coulter and Allen³⁵ reported three infants who had sleep myoclonus that began in the first month of life. The myoclonic jerks were bilateral, repetitive, and located mainly in the forearm and hands. Neurologic examinations and EEG were normal and continued to be normal during follow-up. Coulter and Allen coined the term benign neonatal sleep myoclonus (BNSM) for this phenomenon.

Subsequently, several other small series of cases were published that emphasized the importance of distinguishing BNSM from seizures.^123,143 This condition is probably quite frequent, as suggested in more recent, larger series.^37,42,43

Benign neonatal sleep myoclonus appears in term newborns during the first few weeks of life. One infant born prematurely developed BNSM at 42 weeks of conceptional age,¹²⁰ and the earliest reported onset is in a 5-hour-old newborn.³⁵ The intensity and frequency of the muscle jerks increase up to the third week of life; more subtle myoclonus appearing earlier may go unnoticed. Myoclonic jerks are mainly present during non–rapid eye movement (NREM) sleep; they are less frequent during REM sleep. In some children, BNSM occurs exclusively during NREM sleep.⁴² In most cases, jerks predominately
involve the arms, but the feet, face, axial, and abdominal muscles can also be affected.^{29,37,42,118,123} Myoclonic jerks may be bilateral; localized or multifocal; rhythmic or arrhythmic. The jerks occur frequently in clusters repeating at 1 to 5/sec for several seconds. Clusters of jerks usually recur irregularly in series lasting 20 to 30 minutes⁴² or up to 90 minutes.²⁰ Longer-lasting episodes of BNSM have been mistaken for convulsive status epilepticus.^5,148

Sleep state does not change during the episodes, and arousal always terminates the jerks. Occasionally, BNSM is stimulus sensitive and can be elicited, for instance, by noise.³⁵ Crib rocking can provoke BNSM, which is helpful diagnostically.⁵ Curiously, benzodiazepines may increase the intensity of BNSM.¹²⁰

Benign neonatal sleep myoclonus subsides spontaneously beginning with the second month of life and usually disappears before the sixth month. Ictal EEG is, by definition, normal in BNSM. However, one series reported a higher-than-normal incidence of interictal sharp transients in 4 of 10 newborns.³⁷ Benign neonatal sleep myoclonus may be genetically determined because affected siblings have been reported in two small series of patients, and a history of night jerks in one of the parents has also been found in several cases.^29,143

Although Coulter and Allen³⁵ attributed BNSM to an arousal response, EEG recordings have not demonstrated this. Benign neonatal sleep myoclonus shares some features with nocturnal myoclonus seen in adults,¹⁰⁰ and it may reflect transient immaturity or imbalance of the serotonergic system.¹²³ In our last series of 21 patients, myoclonus disappeared before the age of 7 months, and 2 cases subsequently developed benign myoclonus of early infancy. Prognosis is uniformly good, and no treatment other than reassurance is necessary.⁵⁶

This disorder was described in 1996, and a second series of 13 cases was published in 2001.^153,154 Onset occurs between the first and third month of life, with spontaneous remission 2 to 3 months after onset. These are normal infants presenting diffuse tonic contractions with extension of the four limbs, apnea and cyanosis, without loss of consciousness, lasting 3 to 10 seconds. Seizures occur only during wakefulness and with the child being held in an upright position. Because tactile stimulation often triggers the episodes, reflex myoclonic seizures come to mind. Normal interictal and ictal EEGs help to rule out brief tonic or myoclonic seizures and infantile spasms. The main differential diagnosis is with benign myoclonus of early infancy.

In 1976, Fejerman⁵⁴ reported 10 infants with recurrent spells that resembled infantile spasms but in whom neurologic status, EEG, and outcomes were normal, allowing clear differentiation from West syndrome. Additional cases were added in subsequent reports.^{16,30,55,56,67,74,75,98,101,102} Fejerman termed their conditions benign myoclonus of early infancy (BMEI).

We have now followed a total of 41 patients (26 male, 15 female) for 2 to 27 years.⁶⁰ In general, a diagnosis of West syndrome had been made, even with normal EEGs¹³⁷; the correct diagnosis was established only in retrospect. Infants came to attention because of repeated jerks of the neck or upper limb muscles, causing abrupt flexion or rotation of the head with extension and abduction of the arms. Movements are often described as shuddering of the head and shoulders. In a minority of cases, only the arms are involved; sometimes, shuddering movements alternate with unequivocal myoclonic jerks. Occasionally, the only feature is symmetric or asymmetric extension of one or both arms with abduction, head rotation, or head drops. The jerks may be isolated or repeat in a series. They typically occur multiple times a day.

Consciousness is not affected, even when the myoclonic jerks last as long as 30 minutes. Benign myoclonus of early infancy occurs only rarely during sleep.⁴⁵ Although feeding may trigger attacks, this does not justify a separate designation, as has been proposed.⁵³ Benign myoclonus of early infancy is not related to anger or frustration. Myoclonus appears between 1 and 12 months of age, with onset in 90% of cases between 3 and 9 months. This overlap with the peak occurrence of infantile spasms further confounds diagnosis (Table 2) (see Chapter 229). Furthermore, we reported 4 infants who were normal until onset of typical infantile spasms without hypsarrythmia or other abnormalities in their EEGs. Thus, ictal EEG is very important in differential diagnosis.²⁷

Neurologic examination is always normal, as are interictal and ictal EEG recordings. There are no other laboratory
abnormalities, and behavioral and neurologic development is normal.

Symptoms, once present, may increase for several weeks, but they gradually subside over a few months. Most children are free of attacks by the end of their second year and always by the end of their third. Other conditions that used to be considered in the differential diagnosis of BMEI are listed in Table 3.^{2,49,56,57,133}

Some authors consider shuddering attacks to be different from benign myoclonus of early infancy, and they were initially described as associated with essential tremor.^82,87,150 In fact, I consider that the syndrome named BMEI includes always normal infants who start with fits frequently repeated in bursts, consisting in myoclonic jerks, brief tonic contractions, or shuddering episodes. We do not know the neurochemical basis of this disorder, but we do know that it is a transient phenomenon that may have different neurophysiologic expressions. This interpretation allows the inclusion of some neurophysiologic findings within the spectrum of BMEI.¹⁰⁸

Transient immaturity or imbalance of the serotonergic system might underlie various myoclonic phenomena, as has been postulated for BNSM,¹²³ and which could be genetic, thus accounting for occasional familial cases of both BNSM and BMEI.^29,68 We have seen one child with both BNSM and BMEI. Defining pharmacologic subgroups of the various myoclonic disorders is important, not only for understanding the clinical and neurophysiologic heterogeneity of myoclonus,¹¹² but also for providing rational pharmacotherapy. The areas of brain involved in BMEI are unknown.

This syndrome is characterized by bouts of tonic upward eye deviation accompanied by ataxia.¹⁰⁶ Age of onset is between 6 and 24 months, and children are otherwise healthy. The episodes of upward eye deviation occur in clusters every 2 to 8 seconds over a period of several minutes.^37,51 During episodes, consciousness is preserved, and attempts visually to track objects downward produce vertical nystagmus. Affected children usually exhibit a compensatory tilt of the head with the chin down. Symptoms occur only during wakefulness. During attacks, EEG recordings are normal. The frequency of the episodes gradually declines, and attacks disappear after 1 to 2 years. Even when the episodes disappear, patients may show slow motor development.¹²⁷ Three cases have shown a familial incidence.²³ The brain of one child who died accidentally was normal.¹⁰⁶

Some children have benefited from treatment with dopa, and this has suggested a possible relationship to Segawa syndrome.^23,40

Episodes of paroxysmal dystonia beginning at 3 to 5 months of age were reported in a group of nine infants. The attacks, which occurred only when the children were awake, were characterized by opisthotonus, symmetric or asymmetric increased tone in the arms with extreme pronation of the wrists, and preserved consciousness. Attacks usually lasted several minutes and occurred from once per month to several per day. Sometimes,
however, episodes increased in duration to several hours or even days. In all cases, attacks disappeared in the second year of life.¹⁰

A similar picture had been described earlier as “benign idiopathic dystonia with onset in the first year of life.” These infants had dystonic postures, mainly of the hands, lasting seconds to minutes and associated with intention tremor of the arms.^38,41,157

Of all of these paroxysmal nonepileptic neurologic disorders, benign paroxysmal torticollis probably least resembles epileptic seizures. Benign paroxysmal torticollis is characterized by recurrent episodes of cervical dystonia lasting a few hours or days. Attacks are frequently associated with vomiting, pallor, irritability, and, sometimes, drowsiness or ataxia.^33,39

Onset occurs in the first year of life, and attacks repeat several times per year. Over time, they decrease in severity and disappear after a few years. Electroencephalography is always normal. Gastroesophageal reflux, brainstem/cerebellar, or vestibular dysfunction should be excluded. The benignity of the condition has been emphasized in new series of 22 and 11 cases each.^46,6 In another report, two of the four cases came from a kindred with familial hemiplegic migraine linked to CACNA1A mutation.⁷³ Benign paroxysmal torticollis is now regarded as a migraine equivalent.

Adverse reactions or intolerance to various exogenous agents can include dystonic posturing, shuddering, tonic postures, and myoclonic jerks. Drug ingestion should be sought by careful inquiry, and blood or urine drug screens may be necessary. Metoclopramide,⁷¹ carbamazepine,¹¹⁰ phenothiazines (especially prochlorperazine), droperidol, and trimeprazine¹³⁸ have all been implicated. Intolerance to food additives has also been incriminated as a cause of shuddering attacks.¹²¹

These represent a group of repetitive motor behaviors in children that are typically rhythmic and persistent. They are usually not paroxysmal and, therefore, rarely suggest epilepsy.

Head banging and head turning are common in normal infants at the time of going to bed, especially during the first year of life. Occasionally, children seem withdrawn and inattentive, which may suggest a seizure with automatisms. In some children, the rhythmic head movements persist during sleep (jactatio capitis nocturna). Generally, such movements disappear before 3 years of age, and neurologic development is normal.¹ Similar, often exaggerated motor behaviors occur during wakefulness in children with autistic behavior and mental retardation. A unique case of a parasomnia with rhythmic head and body movements and with tongue biting was reported in a 2-year-old girl.¹⁴⁹