Febrile Seizures: New Concepts and Consequences

By Tallie Z. Baram

Febrile seizures are the most common seizures in infants and children worldwide, This fact provides strong impetus to study and understand them and their consequences, and consider their treatment. These topics were the focus of the first edition of this book.

The 20 years since the publication of this first edition have witnessed an explosion of new information about febrile seizures, meriting this new edition. Key advances have been made in the genetics and neurobiological underpinnings of febrile seizures and especially the very long fever-related seizures called febrile status epilepticus. The role of neuroinflammatory factors in the emergence of these seizures and their consequences, the demonstration of unique clinical and neuroradiological aspects of febrile status epilepticus, and the prospect of predictive (bio)markers to identify and characterize cognitive and epilepsy outcomes are exciting and important. In this edition, the authors and editors tackle these developments in chapters addressing the questions of parents, physicians, allied health care professionals and basic and translational scientists.

• Reviews all aspects of febrile seizures, including epidemiology, neurobiology and treatment
• Discusses novel and newly discovered information based on up-to the minute methods
• Provides an engaging style that is accessible to clinicians, researchers and educated parents

• Language: English
• Publisher: Academic Press
• Release date: Sep 20, 2022
• ISBN: 9780323984188

Book preview

Part I

Epidemiology of febrile seizures

Chapter 1: The incidence and prevalence of febrile seizures and febrile status epilepticus

Rachel Penna; Carl E. Stafstromb    a Division of Pediatric Neurology, The Johns Hopkins University School of Medicine, Baltimore, MD, United States

b John M. Freeman Pediatric Epilepsy Center, Division of Pediatric Neurology, The Johns Hopkins University School of Medicine, Baltimore, MD, United States

Abstract

Epidemiological studies have made substantial contributions to our understanding of the frequency, natural history, and prognosis of febrile seizures. Using a wide variety of methods for case ascertainment, epidemiological studies have consistently estimated that febrile seizures occur in 2%–5% of children under 5 years of age, though there are geographical, seasonal, and sex-related differences in incidence and prevalence. In certain populations (e.g., Japan, Guam), febrile seizures may be even more common (8%–14%). Reasons for these discrepancies remain unexplained, but the susceptibility to febrile seizures likely depends on genetic predisposition, environmental factors, or both. This chapter critically reviews the studies that have led to the widely accepted febrile seizure occurrence rates (incidence, prevalence) and examines the assumptions, methods, and confounding factors underlying those studies.

Keywords

Febrile seizure; Febrile status epilepticus; Children; Epilepsy; Epidemiology

Chapter outline

Introduction

What is a febrile seizure?

Historical perspective

Definitions of febrile seizures

Determining febrile seizure incidence and prevalence

Definitions of epidemiological terms

Study designs to determine incidence and prevalence

Case ascertainment: Was it a febrile seizure?

How common are febrile seizures?

Key studies

Geographical variation in febrile seizures

Seasonal variation in febrile seizures

Febrile seizures are slightly more common in boys

Why do we need epidemiological data?

References

Introduction

Nearly every article or text written about febrile seizures contains a statement along the lines of febrile seizures are the most common type of seizure in childhood, occurring in 2–5% of children [1]. How is this percentage derived, and from what population(s)? Is there a geographical, seasonal, or sex variation in febrile seizure rates? Has the occurrence of febrile seizures changed over time, as genetic diagnosis has become more sophisticated [2], or as different management approaches have evolved? Does the incidence differ according to febrile seizure type (e.g., simple vs. complex)? How can such data help to understand the pathogenesis of febrile seizures and arrive at a consensus for managing them? These are some of the questions considered in this chapter, as a prelude to more detailed discussions of the genetics, natural history, prognosis, and pathophysiology of febrile seizures in subsequent chapters. First, febrile seizure definitions, past and present, are discussed. Next, epidemiological concepts and methods for measuring febrile seizure incidence and prevalence are reviewed. Finally, some pivotal epidemiological studies that have contributed to our current understanding of febrile seizure frequency are discussed.

What is a febrile seizure?

Historical perspective

Over the past several decades, there has been an evolution in what the term febrile seizure connotes. Some series, especially prior to about 1970, did not exclude seizures precipitated by fever that may have been accompanied by an underlying neurological disturbance such as meningitis, encephalitis, or toxic encephalopathy [3,4]. The prognosis of febrile seizures in the early literature was fairly pessimistic, due to the inclusion of symptomatic causes of seizure other than a fever and to patient selection bias [5]. In addition, early studies were predominantly performed at tertiary care facilities and selected the more severe cases [6]. The modern viewpoint is that the majority of febrile seizures, especially simple ones, have a benign outcome, with no lasting neurological sequelae.

The consensus that febrile seizures do not constitute a form of epilepsy was an important conceptual advance with relevance to the consideration of febrile seizure incidence and prevalence. The distinction between febrile seizures and epilepsy has been fairly recent; it was formerly thought that febrile seizures represented epilepsy unmasked by fever and that they were a frequent harbinger of afebrile seizures (epilepsy) [7–9]. Peterman stated emphatically that a febrile convulsion does not occur in a normal child [10]. Bridge (quoted in [11]) [11] stated, in 1949, There is no good reason for considering febrile convulsions as a clinical entity distinct from epilepsy. In reality, both belong in a single group, best described with the name of convulsive disorders. The differences are not of a fundamental nature but only of type and degree [12]. The current definition of epilepsy as unprovoked, recurrent seizures excludes febrile seizures, which are (typically) provoked by fever. Of course, some children experience a febrile seizure as the first manifestation of what will subsequently emerge as epilepsy, and seizure with fever is the canonical presentation of Dravet syndrome (severe myoclonic epilepsy of infancy) [13]. Yet, in the typical child with a febrile seizure, it is not possible to predict with certainty whether that child will develop afebrile seizures [14].

Large epidemiological studies in the 1970s were pivotal in differentiating febrile seizures from epilepsy [3,15–17]. The natural history of febrile seizures is quite different from that of epilepsy, as supported by several observations [18] (Table 1). The risk of developing epilepsy after febrile seizures is small. There are different risk factors for developing febrile seizures vs. epilepsy. About one-third of children with febrile seizures will experience another febrile seizure in a subsequent febrile illness, whereas a maximum of 2%–4% will later develop afebrile seizures (epilepsy) [3,16,19–23]. The major predictors of recurrent febrile seizures are as follows: occurrence before 1 year of age, a positive family history of febrile seizures (not a family history of epilepsy), and a low degree of fever [24]. On the other hand, the risk factors for epilepsy in children with febrile seizures are as follows: complex febrile seizures, a family history of epilepsy, and neurologic impairment prior to the febrile seizures [22]. One study did find that an initial complex febrile seizure increased the risk for febrile seizure recurrence [25]. Notably, prevention of recurrent febrile seizures does not appear to alter the risk of developing afebrile seizures. Several randomized studies of febrile seizure prevention showed that although febrile seizure recurrence could be reduced, the development of later afebrile seizures was not altered by treatment [14,26–28].

Table 1

The controversy as to whether febrile seizures initiate a pathophysiological cascade that ultimately results in mesial temporal sclerosis and hence temporal lobe epilepsy (TLE) has been unresolved for decades [29,30]. A disproportionate number of patients (up to 20%) with TLE had febrile seizures as young children, especially prolonged ones or febrile status epilepticus [31]. However, prospective series cite only a minority of children with febrile seizures going on to develop epilepsy [15,16,32,33], and both types of studies suggest that the risk for epilepsy is confined almost exclusively to prolonged and complex febrile seizures and febrile status epilepticus, a finding supported also by work in experimental models [34–36].

Definitions of febrile seizures

National Institutes of Health, International League Against Epilepsy, and American Academy of Pediatrics Definitions

According to the International League Against Epilepsy (ILAE) classification of the epilepsies, febrile seizures are an acute, symptomatic type, that is, a special, situation-related seizure [37]. Because they are always evoked by a specific precipitant (fever), febrile seizures cannot be considered a form of epilepsy. Nevertheless, febrile seizures do constitute a syndrome because they fulfill several characteristics that are similar among affected children:

1.Febrile seizures generally occur within a restricted age range.

2.The majority of febrile seizures occur in children who are neurologically normal and continue to develop normally after febrile seizures.

3.Febrile seizures are not associated with a structural or developmental anomaly of brain, although the existence of such pathology may enhance the susceptibility to febrile seizures [38].

Two operational definitions of febrile seizures have been published, one from a National Institutes of Health (NIH) Consensus Conference [39,40], and the other from the ILAE [41]. A third, more recent definition from the American Academy of Pediatrics [42,43], has been used in recent epidemiological assessments and treatment guidelines [44]. These definitions are compared in Table 2. In light of the huge disparities in older epidemiological studies of febrile seizures, the 1980 NIH conference represented a major advance, forming a coherent definition that has been used subsequently in many epidemiological and therapeutic investigations [20,45–47]. The ILAE proposed guidelines for epidemiological studies of epilepsy in general, as an attempt to standardize methods of case ascertainment, diagnostic accuracy, and seizure classification, and these recommendations are relevant to studies of febrile seizures as well. The recently updated epilepsy classification scheme highlights the recognition that Dravet syndrome can present with a seizure in presence of fever [48].

Table 2

Although the NIH, ILAE, and AAP definitions are similar, their differences are worth noting. The lower age limit for febrile seizures is 1 month in the ILAE definition, 3 months in the NIH definition, and 6 months in the AAP definition. The NIH guideline is made purposefully flexible by use of the phrase usually occurs. Each of these lower age limits has been employed in epidemiological studies of febrile seizures (1 month [15,16,24,49,50], 3 months [20,25,51–53], and 6 months [23,54,55]). Some studies do not state a lower age limit. All definitions exclude children with prior afebrile seizures and those with seizure due to an intracranial infection or other specific cause, but none excludes children with prior neurological impairment. The ILAE definition subdivides febrile seizures into those with and without prior neonatal seizures. Seizures with fever occurring during the second and third months of life, included in the ILAE definition, might not be included under the NIH definition. Practically speaking, children in this young age range probably account for a very small proportion of cases [56] (Fig. 1). The specific age during the first few months of life when the maturing brain first expresses an increased susceptibility to fever-induced seizures is not known precisely and may vary somewhat between children; this variability theoretically could impact epidemiological data. For example, the 6-month lower age limit in the AAP definition would miss early onset cases.

Fig. 1

Fig. 1 Age-specific incidence of febrile seizures among children in Japan; Rochester, Minnesota; and Oakland, California. (From Hauser WA. The prevalence and incidence of convulsive disorders in children. Epilepsia 1994;35 (Suppl. 2):S1-S6. Reprinted by permission of Blackwell Science, Inc.)

Another notable difference is the lack of an upper age limit in the ILAE definition. The majority of febrile seizures occur between 6 months and 3 years of age, with the peak incidence at about 18 months (Fig. 1) [3,15,19,21,45,57,58]. The observation that these data comprise a bell-shaped curve, regardless of the population studied, attests to the unique age specificity of the brain’s sensitivity to fever. Only about 6%–15% of first febrile seizures occur after 4 years of age, and onset after 6 or 7 years of age is distinctly unusual and should call the diagnosis into question.

None of the definitions includes a specific criterion as to what temperature defines fever. A temperature of at least 38.4°C (101°F) would probably be accepted by most authorities and has been utilized in many epidemiological studies [47]. Some researchers have accepted lower values (38°C) [25,58]. The site of temperature determination is not always mentioned. Other rigorous studies have differentiated axillary (37.8°C) from rectal (38.3°C) temperatures and have required that these be documented by emergency room personnel [24]. Despite the common belief that the rate of temperature rise is more important than the ultimate temperature achieved, there is little data supporting that view [59,60]. Many febrile seizures (at least half) occur early in the course of a febrile illness, especially within the first 24 h, and can even be the presenting sign of the febrile illness [18,24,61]. In clinical practice, it is difficult to establish the exact temperature just prior to or at the onset of a febrile seizure. Such a determination would ordinarily be in the hands of the parent or other caretaker. In most cases, a temperature is not recorded by the parent at the time of the seizure, and by the time the child’s temperature is recorded at the physician’s office or emergency room, the seizure is likely to be over and the information may be of limited relevance. Despite the lack of a uniform temperature criterion, several studies have shown that the height of the fever is related to febrile seizure recurrence [60,62].

An explicit definition of seizure is not specified in either NIH, ILAE, or AAP definition. Some authorities emphasize the presence of convulsive activity [4,63]; indeed, febrile seizures are also known as febrile convulsions. But is convulsive activity necessary? The lack of tonic-clonic rhythmic motor activity certainly makes the diagnosis of febrile seizure less secure, but clinicians routinely diagnose febrile seizures in children who present with limpness, altered consciousness, apnea, or nonconvulsive activity suggestive of focal unaware symptomatology. In fact, the NIH definition uses the vague term event rather than seizure, perhaps to emphasize that some febrile seizures may not involve convulsive activity. Other clinical phenomena in the differential diagnosis of febrile seizures are mentioned in the section on Case Ascertainment below. Inclusion of nonconvulsive activity under the febrile seizure umbrella could alter epidemiological conclusions.

Simple and complex febrile seizures

Febrile seizures are typically divided into two types, simple and complex. A complex febrile seizure has one or more of the following features [41,64]: focal onset or focal features during the seizure, prolonged duration (>  10 min [21,24] or >  15 min [16,65]), or recurrence of febrile seizures within 24 h of the first episode [15,21]. Some authors use a phrase such as within the same febrile illness instead of within 24 hours [63,66]. A simple febrile seizure consists of <  10 or 15 min of seizure activity, resolving spontaneously, in the context of a febrile illness, without focal features or recurrence during the subsequent 24 h.

Approximately 20%–30% of febrile seizures are complex [15,20,25,45,64,65,67]. (The term complicated febrile seizure may be more accurate than complex, to avoid confusion with partial complex seizures, but most literature continues to use the term complex febrile seizure.) Early studies that showed a much higher percentage of complex febrile seizures (e.g., 62% in [68]) were comprised of hospitalized patients, representing the more severe end of the spectrum. In the National Collaborative Perinatal Project (NCPP), a cohort of 54,000 children was followed from before birth until 7 years of age [15]. In this study, 1706 children with febrile seizures were identified. Focal onset was present in 4%, prolonged duration (>  15 min) in 8%, and recurrent febrile seizures within 24 h of the first one in 15%–20%. Complex febrile seizures are associated with a higher risk of afebrile seizures but not of febrile seizure recurrence [18,20]. Because complex febrile seizures are associated with a higher risk of subsequent epilepsy, it is important to establish whether complex features are present when evaluating a child with a febrile seizure. Of the complex features, recurrence within 24 h is perhaps the easiest to document. The presence of focal features at onset can often be overlooked by the observer. Finally, as discussed in the section on Case Ascertainment, estimation of seizure duration may be quite inaccurate. As discussed in depth elsewhere in this volume, the two subtypes of febrile seizures may form biologically distinct conditions with different risks for future seizures and neurologic deficits.

Febrile status epilepticus

Febrile status epilepticus (FSE) is a term defined by the ILAE as a febrile seizure lasting at least 30 min [41]. Other studies have interchangeably used the term prolonged febrile seizure to denote a febrile seizure duration of at least 15 min [20,65,69]. While FSE accounts for 5%–9% of all febrile seizures [65,70,71], it represents about 25% of all pediatric status epilepticus and over two-thirds of the cases in the second year of life. A retrospective study of hospitalized children from Nepal showed that 4.6% of all children admitted for febrile seizures had FSE [72].

A multicenter study led by Shinnar has been fundamental to understanding the characteristics of FSE [73]. Out of 180 cases of children with FSE, 123 (68%) were younger than 2 years of age and 128 (96%) were younger than 5 years of age, similar to the age of onset of children with brief febrile seizures. Approximately one-third of the seizures had focal onset. Durations of FSE were 30–59 min in 58%, from 1 to 2 h in 24%, and over 2 h in 18% of cases [73]. The short-term morbidity and mortality of FSE was extremely low, consistent with prior studies of febrile seizures [3,15,16,21,66,74].

A growing area of research is the relationship between FSE with hippocampal injury, mesial temporal sclerosis, and temporal lobe epilepsy. The multicenter FEBSTAT study (Consequences of Prolonged Febrile Seizures in Children) has prospectively followed 199 children with febrile status epilepticus and has elucidated long-term outcomes of FSE [75,76]. Approximately 12% of cases had MRI evidence of increased T2 signal in the hippocampus following FSE compared with no patients in the control group with brief febrile seizures [77]. After 1 year, the children with febrile status epilepticus had smaller hippocampi than controls with simple febrile seizures, and those with initial T2 hyperintensity following FSE had evolved to hippocampal sclerosis on MRI [78]. Additional follow-up studies are needed to further elucidate the relationship between FSE with hippocampal injury and temporal lobe epilepsy, as detailed in subsequent chapters.

In summary, epidemiological studies vary in their operational definitions of febrile seizures, age limits employed, methods for determining whether they are simple vs. complex, seizure duration, and other factors. Such variations need to be considered when comparing studies.

Determining febrile seizure incidence and prevalence

Definitions of epidemiological terms

Much of our knowledge about the risk factors, natural history, and prognosis of febrile seizures comes from epidemiological studies. However, epidemiological conclusions are only as reliable as the initial data obtained. To analyze the literature on febrile seizure frequency, certain epidemiological terms must be defined (Table 3) [79–81]. Unfortunately, incidence and prevalence are not always used correctly in the literature, and the terms are sometimes used interchangeably.

Table 3

Incidence (or incidence rate) is the number of new cases occurring in a defined population over a specified period of time. For example, febrile seizure incidence is often denoted as the number of new cases per 1000 persons in a population, per year. Incidence can only be determined from longitudinal studies. Prevalence is the proportion of individuals in a population that has ever had the disorder, determined at a specific time point. Prevalence can be obtained from cross-sectional surveys. For example, prevalence is often specified as the number of children with a history of febrile seizures on a given date that a population survey is performed. Prevalence is dependent on both the incidence of the disorder and its average duration. Because the number of cases accumulates over time, prevalence can be high even if incidence is low, and knowledge of prevalence does not always lead to accurate incidence statistics. Although both incidence and prevalence can be adjusted for age, prevalence is not as useful in an age-specific syndrome such as febrile seizures. Febrile seizure incidence should decrease with age and become zero after 6–7 years of age (Fig. 1).

Cumulative incidence is the summation of age-specific incidence rates, that is, expected risk of developing febrile seizures by a specific age. Therefore, this term is probably the most appropriate one to compare febrile seizure occurrence between populations, because nearly all cases are expected to occur by about 5 years of age. Cumulative incidence should approach the lifetime prevalence that is the proportion of the population that has ever had a febrile seizure. This point is illustrated well by van den Berg and Yerushalmy [3], who showed that the cumulative incidence reaches a plateau by about 4 years of age. Annual incidence refers to the incidence only within the year studied; this value should be summated over the years of febrile seizure susceptibility to arrive at a rate approximating the cumulative incidence. Finally, the first attendance rate[82] is defined as the number of new cases in the population at risk per year, including both new cases during the study period and those who had their first febrile seizure prior to the study period but were diagnosed during the study period.

Study designs to determine incidence and prevalence

The wide variety of study designs may strongly impact incidence and prevalence figures. To determine those rates, a number of techniques have been used. The goal of any method is to determine the number of cases open (numerator) and the total population from which the cases are drawn (denominator). The two main types of study designs are clinic-based and population-based. Clinic-based studies are often performed in a specialized clinic, hospital setting, or other health care facility. They have the advantage of providing a readymade set of patients with detailed clinical information available. However, the type of patient that seeks care or has access to such a facility may skew the results. In general, these would be the most severely affected patients, with the poorest outcomes. Clinic-based studies of febrile seizure occurrence are rarely performed anymore.

Many studies have used larger populations to investigate febrile seizure incidence and prevalence, using case-finding methods such as medical record reviews, mailed questionnaires, telephone interviews, and door-to-door surveys (Table 4). Although these techniques are useful to investigate a much larger population base and thus better reflect the spectrum of disease severity and frequency, they are also subject to methodological biases. Such large-scale surveys are ordinarily carried out by personnel with limited medical training. Home visits are expensive and time consuming. Medical records are subject to some of the same biases discussed above, with documentation dependent on the training, time, and interest of the personnel collecting the data.

Table 4