FCs are the single most common problem in pediatric neurology. Although the definitions and methods of case ascertainment have varied, that 2% to 5% of all children will experience at least one convulsion
with febrile illness before 5 years of age is widely accepted (Fukuyama, 1991; Forsgren et al., 1990b; Annegers et al., 1987; Verity et al., 1985a; Leviton and Cowan, 1982; Lennox-Buchtal, 1973). Higher rates have been reported in some nonoccidental countries, including rates of 7% in Japan and of up to 14% in the Marianna Islands (Tsuboi, 1984, 1986; Mathai et al., 1968). These high rates may be explained by several factors. In developing countries, some common infections of childhood (e.g., measles) may occur earlier, on an average, than in industrialized countries, thus taking place during the period of maximum susceptibility to FCs; the number and/or severity of febrile diseases may therefore be greater.

Living conditions (e.g., the whole family living in a single room) may result in improved case identification. Other unknown factors, whether genetic or environmental, also may play a role. FCs are more common in males (Hauser, 1981; Lennox-Buchtal, 1973; Millichap, 1968), with the sex ratio varying between 1.4 to 1 and 1.2 to 1. This male excess may be the result of an excess of one-sex (male) sibships (Ounsted et al., 1966).

The main risk factors for FCs include age, fever, and genetic predisposition. Other factors of lesser importance may be a history of abnormal prenatal and perinatal events that may facilitate the occurrence of FCs or influence their clinical expression and outcome (Forsgren et al., 1991; Wallace, 1976a, 1976b Chevrie and Aicardi, 1975). In one study, heavy smoking during pregnancy was associated with a higher rate of febrile seizures in offspring (Nelson and Ellenberg, 1982, 1990) and was perhaps related to the increased frequency of respiratory illness in young children exposed to passive smoking, but this relationship was not found in another study (Forsgren et al., 1990b).

The initial development of the child has a definite impact on the expression and the outcome of febrile seizures. Wallace (1972, 1976b) considered FCs a possible indication of lifelong developmental defects of prenatal or perinatal origin.

The studies of Wallace (1972, 1976b) and those of Nelson and Ellenberg (1982, 1990) have found that previous developmental abnormalities were associated with complex febrile seizures and that these had a less favorable outlook. Nelson and Ellenberg (1981b) found an extraordinarily high incidence of suspect or abnormal development before the occurrence of FCs on at least one examination in the large National Collaborative Perinatal Project (NCPP). In this study, 22% of their patients had such a history. However, the rate of abnormal development in nonconvulsive children is not stated, although that it was also quite high, possibly on the order of 12%, has been indicated by Nelson and Ellenberg (1976). Thus, the population of children with FCs likely was somewhat biased by the presence of children considered worrisome or abnormal on early examination when compared with a normal population, but the precision of the measurement limits the value of such evaluations.

All febrile seizures are either tonic-clonic or possibly hypotonic (Gastaut et al., 1962). They never manifest as myoclonic seizures, spasms, or nonconvulsive attacks. The great majority are bilateral clonic or tonic-clonic attacks of short duration (less than 15 minutes). These are termed simple FCs. These seizures are followed by very brief postictal manifestations but no others; their virtual absence is of great diagnostic significance. Complex febrile seizures are those of long duration (15 minutes or more), those with focal features (most commonly unilateral), or those that recur two or more times within a single illness episode. Up to one-third of febrile seizures may have one or more of these complex features. Seizures of long duration were noted in 18% of the children in the Gentofte series (Frantzen et al., 1968), in 16% in an old series by Herlitz (1941), and in 35% in studies by Wallace (1974, 1976a, 1988). In the large population study of Nelson and Ellenberg (1978), 4.3% of the FC attacks lasted longer than 30 minutes and 7.6% longer than 15 minutes; 4% had a focal onset. Some of the unilateral seizures may be followed by a Todd hemiplegia that usually lasts a few hours but that may persist for up to several days. The incidence of Todd hemiplegia is probably in the range of 0.4% of all cases of FC (Nelson and Ellenberg, 1978). Seizures lasting 30 minutes or more qualify as status epilepticus. Rarely, they can constitute the initial stage of the hemiconvulsion-hemiplegia syndrome or acquired postconvulsive hemiplegia (see Chapter 10). However, the outcome is favorable in most cases of febrile status (Shinnar et al., 1992; Maytal and Shinnar, 1990). Such patients need emergency treatment. Most long-lasting convulsions (approximately 75%) are the initial seizure (Nelson and Ellenberg, 1978; Aicardi and Chevrie, 1970, 1975), an important finding that indicates the impossibility of prevention.

The occurrence of more than one seizure in 24 hours during the same febrile episode is not uncommon; it amounted to 16% of the cases in the Collaborative Childhood Perinatal Project series (Nelson and Ellenberg, 1976). Repeated seizures are also seen in the initial attacks in 6% of cases. They are usually brief convulsions, and they generally occur within 2 to 4 hours after the onset of fever.

Complex seizures more commonly occur in children with previously abnormal development or neurologic findings than do simple seizures. Aicardi and Chevrie (1983) showed that children with unilateral FCs, most of which were also of long duration, had a lower incidence of a positive family history of FCs than did those with brief bilateral seizures (18% versus 35%) and a higher incidence of abnormal and perinatal antecedents, thus suggesting the possible role of acquired, possibly lesional factors in their determination. Their prognosis, however, is favorable, although they may carry a slightly increased risk of febrile recurrences (Offringa et al., 1994).

Paroxysmal EEG abnormalities were found in 35% to 45% of patients for whom serial EEGs were obtained before the age of 5 years (Doose et al., 1983; Lennox-Buchtal, 1973). Sofijanov et al. (1992) found that the initial EEG of 676 children with FCs contained
paroxysmal abnormalities in 22%. In all series, spike-waves at 3 Hz or more were most common, with focal spikes being second in frequency (about 10% of cases) (Lennox-Buchtal, 1973).

Generalized spike-waves should be distinguished from the so-called hypnagogic bursts that are a normal finding in children in the age bracket of FCs and that therefore are often found in these patients (Alvarez et al., 1983). The presence of paroxysmal EEG abnormalities is correlated with age. Paroxysmal anomalies are rare before the age of 1 year, but they are found in 51% of patients who had an initial EEG recording after 4 years of age (Sofijanov et al., 1992). Paroxysms are also significantly more common in children who had complex febrile seizures lasting more than 15 minutes, but they are related neither to the birthweight, gender, the nature of the febrile illness, the family history of convulsions, or epilepsy nor to the occurrence of multiple seizures in the same 24-hour episode. The number of previous febrile seizures is strongly related to the occurrence of EEG anomalies. These were found in 18% of children who had a single seizure and in 63% of those who had four or more seizures (Sofijanov et al., 1992). In the same study, EEG paroxysms were also more common in children with abnormal neurologic signs antecedent to the convulsions and with focal seizures, but the association was not as close as that of age and number of previous convulsions.

Most prospective studies have found no correlation between the presence of EEG paroxysms and the later emergence of nonfebrile seizures (Lennox-Buchtal, 1973; Frantzen et al., 1968). Millichap and Colliver (1991) did, however, find the incidence of paroxysmal EEG abnormalities to be five times greater in children who developed epilepsy than in those who did not. Even so, the EEG is of little, if any, practical value for the prediction of recurrences or epilepsy in the individual patient; paroxysms are found in a high proportion of the older patients who are least likely to develop severe or recurrent seizures of epilepsy, whereas they never appear on the EEGs of the young infants who are at maximum risk for recurrences, status epilepticus, and epilepsy. Biparietal theta rhythms were present during wakefulness in 54% of their patients, spike-wave activity in the resting record in 49%, and photosensitivity in 42%. The last two patterns were strongly age dependent.

The diagnosis of FCs is not always easy. Clearly, seizures associated with CNS infections, a topic discussed later in this chapter (see “Investigations in Children with Febrile Convulsions”), must not be diagnosed as FCs. Convulsions associated with fever may also be observed with toxic exposure, septic embolization, hemolytic-uremic syndrome, and other acute encephalopathies (Nelson and Ellenberg, 1983). Biotinidase deficiency may be responsible for both seizures and repeated infections (Aicardi, 1992a), and this condition is important to recognize because it is lethal condition if left untreated. Malaria and other parasitic diseases are important in many parts of the world, and these require specific therapy (Newton and Warrell, 1998). Severe FCs may raise the problem of the acute encephalopathies of obscure origin. Lyon et al. (1961) has suggested that some of the cases of acute encephalopathy in childhood are nothing more than severe and complicated cases of FCs. Whether such cases should be separated from FCs remains uncertain, as discussed in the introduction to this chapter.

Gastaut (1974) and Stephenson (1983, 1976, 1990) emphasized the fact that seizures occurring with fever were not necessarily of an epileptic nature. Syncopes, or reflex anoxic seizures, also can be triggered by fever, and the distinction between anoxic and epileptic seizures may be difficult (see Chapter 21). Stephenson (1976) divided 100 children who had fits with fever into the following three subgroups on the basis of the clinical description of attacks; 14 children were considered without doubt to have anoxic seizures, 35 had epileptic seizures, and the remaining 51 children could not be definitely classified. An exaggerated oculocardiac reflex was rare in the epileptic group, it was extremely common in the anoxic group, and it occurred with intermediate frequency in the undetermined group. However, an excessive oculocardiac reflex may be found in a child with authentic epileptic seizures, and the findings of Stephenson (1976) are difficult to interpret, even though they certainly suggest that febrile anoxic seizures may be common.