Address correspondence and reprint requests to Prof. Dr. D. Rating at University Children's Hospital Heidelberg, Department of Pediatric Neurology, Im Neuenheimer Feld 150, 69120 Heidelberg, Germany. E-mail: firstname.lastname@example.org
Summary: Purpose: To estimate the incidence rate of epilepsies and epileptic syndromes in German children and adolescents aged 1 month to <15 years, and to provide data on their classification.
Methods: A population-based prospective study was performed between July 1, 1999, and June 30, 2000. All children aged 1 month to <15 years with a newly diagnosed epilepsy or epileptic syndrome were recorded by private pediatricians, EEG laboratories, and the two University Children's Hospitals in the neighboring cities of Heidelberg and Mannheim. The diagnoses were classified according to the International Classification of Epilepsies and Epileptic Syndromes of the International League Against Epilepsy (ILAE).
Results: The total age-adjusted annual incidence rate was 60/100,000 (95% confidence interval, 42–84), with the highest incidence in the first year of life (146/100,000). Focal epilepsies or epileptic syndromes (58%; incidence rate, 35/100,000) were more common than were generalized ones (39%; incidence rate, 24/100,000), and 3% (incidence rate, 2/100,000) of the epilepsies or epileptic syndromes were undetermined. The rate of idiopathic (47%; incidence rate, 29/100,000) and symptomatic or cryptogenic epilepsies (50%; incidence rate, 30/100,000) was equal. No significant difference in incidence between boys and girls was found.
Conclusions: Incidence rates for epilepsy in German children aged 1 month to <15 years are about equal to those of other countries in Europe and North America. In accordance with studies from the United States and from many European countries, incidence was highest in the first year of life, and no difference in the incidence between girls and boys was found. In Germany as throughout Europe, idiopathic generalized epileptic syndromes are more often diagnosed than in the United States.
Epilepsies are the most common neurologic diseases in children and adolescents. The rate of new diagnoses [i.e. the incidence rate (IR)], is of special interest for health system planning and forms the basis for risk calculations.
In Germany, three studies on the incidence of epilepsies in children and adolescents were performed (1–3). Incidence rates were retrospectively assessed by gathering information from hospital charts. Children with a first epileptic seizure were included. Incidence rates ranged from 31/100,000 to 202/100,000 dependent on age and evaluation year.
Other studies from North America (4–8) and Europe (9–23) were designed either as retrospective studies (4–14) or as prospective studies (15–23). With the exception of three studies (18,19,22), epilepsy was diagnosed at the occurrence of the second unprovoked seizure, as proposed by the Guidelines on Epidemiological Studies by the ILAE (24). Incidence rates ranged from 20/100,000 to 375/100,000, dependent on age group, gender, and country. In every study, the highest incidence was reported for the first year of life; after the tenth year, the rate decreased.
In Germany no incidence study with a prospective design has been performed to date, and all three retrospective studies (1–3) did not use the diagnostic criteria of the ILAE Guidelines (24) or Classification (25). Therefore we performed a prospective study to investigate the incidence of epilepsies and epileptic syndromes in childhood using ILAE criteria (25–27).
We collected prospectively data for all children and adolescents aged 1 month to <15 years in the cities of Heidelberg and Mannheim during 1 year (July 1999 to June 2000), who were newly diagnosed with an epilepsy or epileptic syndrome.
Investigational area and population
Heidelberg and Mannheim are neighboring cities in the southern part of Germany with a total population of 448,553 (census of December 31, 1999 by the statistics departments of both cities). The total number of children and adolescents aged 0 to <15 years was 59,647 (Heidelberg: 16,804, 28.2%; Mannheim: 42,845, 71.8%). The proportion of boys (51.2%) was slightly higher than the proportion of girls (48.8%). One fourth of the children and adolescents were not German citizens (24.5%). The age distribution of the population in the investigational area is representative for that age group (0 to <15 years) in Germany (census of December 31, 1998 by the Federal Statistic Office of Germany).
First, all private pediatricians in Mannheim and Heidelberg, including the four EEG laboratories besides the children's hospitals, the attending medical doctor for pediatric neurology (S.K.) at the University Children's Hospital in Mannheim, and a medical doctor (C.M.F.) at the University Children's Hospital in Heidelberg subsequently filled in a questionnaire for each child, after the child was seen because of a suspected seizure. In both children's hospitals, all children with a suspected seizure or attending for an EEG were checked for a new diagnosis of epilepsy. The private pediatricians, including the four EEG laboratories, were contacted on a regular basis (letters, phone calls, personal visits) and cooperated very well. Second, the medical records of all children with a second unprovoked seizure during the assessment period were reviewed. In Germany, almost every single person is covered by health insurance, including children from immigrant families. Most parents who recognize seizures or neurologic disorders in their children will consult a pediatrician immediately. In Germany, as a rule, a child in case of a suspected seizure will first be seen by a pediatrician and then be referred to a pediatric neurologist. In our study, all children seen by a private pediatrician were referred at least once to the two previously mentioned hospitals. Medical records of all children existed at the hospitals in addition to the information from the private pediatricians.
Strictly speaking, our method of assessment provides incidence rates based on the first attendance of children with a suspected epilepsy. However, we assume that the true incidence is only slightly underestimated by the first attendance rate.
The diagnoses of epilepsies and epileptic syndromes were validated by two pediatric neurologists (D.R.,S.K.) by reviewing the history, results of the neurologic examination, EEG, magnetic resonance imaging (MRI), and the laboratory examinations of each case. Awake EEG and/or sleep-deprived EEG were recorded in all cases; MRI was performed in case of a suspected diagnosis of symptomatic or cryptogenic epilepsy. Classifications were made according to the ILAE criteria (25) and based on a constellation of factors including age, seizure descriptions and frequency, EEG, underlying etiology, and diurnal seizure pattern.
Concerning the diagnosis of “other idiopathic generalized epilepsy,” all five children had an epilepsy diagnosed in Germany as “early infantile epilepsy with generalized tonic–clonic seizures,” which is not included in the ILAE classification of epilepsies and epileptic syndromes. This syndrome is characterized by generalized tonic–clonic seizures, initially often as febrile convulsions, beginning between ages 6 months and 4 years. Sometimes, the generalized tonic–clonic seizures are lateralized; myoclonic seizures can follow. In the beginning, the EEG is often normal, changing later to a spike–wave pattern (28,29). This syndrome is comparable to the epilepsy with generalized tonic–clonic seizures in childhood, showing an unfavorable prognosis, as it was described in two studies (30,31).
Inclusion and exclusion criteria
Epilepsy was diagnosed at the time of the second unprovoked seizure. Children with seizures occurring only in the first month of life and children with single or recurrent febrile convulsions were excluded. We also excluded children with a first unprovoked seizure or with recurrent or complicated febrile convulsions who received subsequent antiepileptic drug (AED) treatment. An episode of status epilepticus or several seizures during a 24-h period were considered single events.
Fisher's Exact tests were calculated with SPSS 10.0 for Windows. Exact 95% confidence intervals were calculated using the program STATISTICA (module Power Analysis; Statsoft, Tulsa, OK, U.S.A., version 5.5, 1999) (32).
Annual incidence of epilepsies and epileptic syndromes
Thirty-six children, 10 (27.8%) from Heidelberg and 26 (72.2%) from Mannheim, aged 1 month to <15 years, received a new diagnosis of epilepsy or epileptic syndrome during the 1-year observation period.
The results of the annual incidence in different age groups and different epilepsies are summarized in Table 1, Fig. 1, and Fig. 2. The age-adjusted incidence rate (IR) in the German population was 60.3/100,000 for age 1 month to <15 years. The estimated overall cumulative incidence was 0.9%. No significant difference was found in the raw scores for girls and boys in different age groups (Fisher's Exact test, two-sided, p = 0.58). There was a trend for a higher incidence in the youngest children (1 to <12 months) compared with the older children (one to <15 years; Fisher's Exact test, p = 0.06). No significant difference in raw scores for focal and generalized epilepsies between different age groups was found (Fisher's Exact test, two-sided, p = 0.27). Focal and generalized epilepsies were equally distributed in girls and boys. The diagnoses of different epilepsies and epileptic syndromes are given in Table 2.
Table 1. Raw annual incidence rate for different age groups
Population at risk
CI, confidence interval.
1 mo–<10 yr
1 mo–<15 yr
Table 2. Patients grouped according to the ILAE Classification of Epilepsies and Epileptic Syndromes Classification
No. of cases (%)
IR, 35,4/100,000 (95% confidence interval, 21.9–54.1).
IR, 23.6/100,000 (95% confidence interval 12.9–39.6).
IR, 1.7/100,000 (95% confidence interval 0.0–9.4).
Benign childhood epilepsy with centrotemporal spikes
Childhood epilepsy with occipital paroxysms
Primary reading epilepsy
Chronic progressive epilepsia partialis continua of childhood
Syndromes characterized by seizures with specific mode of precipitation
Temporal, frontal, parietal, and occipital lobe epilepsies
Temporal, frontal, parietal, and occipital lobe epilepsies
2. Generalized epilepsies and epileptic syndromesb
2.1 idiopathic (with age-related onset)
Benign neonatal familial convulsions
Benign neonatal convulsions
Benign myoclonic epilepsy in infancy
Childhood absence epilepsy
Juvenile absence epilepsy
Juvenile myoclonic epilepsy
Epilepsy with generalized tonic-clonic seizures on awakening
Other idiopathic generalized epilepsies not defined above
Epilepsies with seizures precipitated by specific mode of activation
2.2 Cryptogenic or symptomatic (in order of age)
Epilepsy with myoclonic–astatic seizures
Epilepsy with myoclonic absences
2.3.1 Nonspecific etiology
Early myoclonic encephalopathy
Early infantile epileptic encephalopathy with suppression burst
Other symptomatic generalized epilepsies not defined above
2.3.2 specific syndromes
Epilepsies due to specific neurologic diseases
3. Epilepsies undetermined whether focal or generalizedc
3.1 With both focal and generalized seizures
Severe myoclonic epilepsy in infancy
Epilepsy with continuous spike–waves during slow-wave sleep
Acquired epileptic aphasia
Other undetermined epilepsies not defined above
3.2 Without unequivocal focal or generalized features
4. Special syndromes
As shown in Table 2, the etiology of seizures was idiopathic in 17 (47.2%; IR, 28.7/100,000; 95% CI, 16.7–45.9/100,000) and symptomatic or cryptogenic in 18 cases (50.0%; IR, 30.3/100,000; 95% CI, 18.0–47.9/100,000). Idiopathic etiology was found almost twice as often in the generalized epilepsies (9/14; 64.3%) than in the focal epilepsies (8/21; 38.1%), but the difference was not statistically significant (Fisher's Exact test, two-sided, p = 0.18). Epileptic seizures in first-degree relatives were found in eight (22.2%) children. Three of those children were diagnosed with epilepsy with myoclonic astatic seizures, two with benign childhood epilepsy with centrotemporal spikes, one with juvenile absence epilepsy, and one with juvenile myoclonic epilepsy. In the case of one girl from the study group with symptomatic temporal lobe epilepsy, the elder brother had the same epilepsy, and the same unknown mental retardation syndrome as did his sister.
Four (11.1%) children were born with a central nervous system (CNS) malformation, and two (5.6%) had experienced perinatal complications. Five (13.8%) children were severe mentally retarded; two (5.6%) showed developmental delay. Of the severely mentally retarded children, one child had an Angelman syndrome and two had unknown dysmorphic syndromes.
The initial AEDs administered were carbamazepine (CBZ) in 17 (53.1%), valproic acid (VPA) in 13 (40.6%), and sulthiame (SLT) in two (6.3%) cases (33). Four children were not treated. For the focal epilepsies, CBZ was the main medication (14 of 17 treated focal epilepsies, 82.4%). The generalized epilepsies were treated mainly with VPA (n = 12; 85.7%).
In this prospective population based study, we recorded the first attendance rate of newly diagnosed epilepsies and epileptic syndromes in children, as children whose parents or caregivers did not recognize epileptic seizures did not enter the study. However, as mentioned earlier, the health system in Germany is easily accessible, so in case of a suspected seizure, parents are very likely to visit a doctor. Therefore we assume that the incidence rate is only slightly underestimated. However, children with seizures not easily recognized, like absences and complex focal seizures, might not have entered the study.
Annual incidence in different age groups and sex distribution
The results of this first prospective study in Germany on the incidence of epilepsies in children conducted along the ILAE Guidelines for diagnosis and classification (24,25) are not comparable with those of the three other German studies, because they took the first seizure into account, which may result in a 33% overestimation of incidence (5). Our total annual age-adjusted incidence rate of 60.3/100,000 for children aged 1 month to <15 years is similar to that of other studies from North America and Europe (see Table 3; i.e., countries with comparable health systems and standard of living). Low incidence rates are found in the Canadian (6) and one U.K. study (16). Both studies reported an incidence rate of 41–43 for children aged 1 month to 14 years. This might be due to retrospective design in the Canadian study. For the U.K. study, it might be due to incomplete case-ascertainment methods, related to their use of screening questionnaires (26,27) or to dropout. It may be concluded that the incidence rate of epilepsies and epileptic syndromes for children aged 1 month to <15 years in western countries is about equal and amounts to 60/100,000.
Table 3. Gender-specific incidence rates for different age groups in comparable studies
1–12 mo rate/100,000 (number of cases)
1–<5 yr rate/100,000 (number of cases)
5–<10 yr rate/100,000 (number of cases)
10–<15 yr rate/100,000 (number of cases)
1 mo–<15 yr rate/100,000 (number of cases)
Rochester, NY, U.S.A. (4,5)
Rochester, NY, U.S.A. (8)
Nova Scotia, Canada (6)
Rural Iceland (21)
Västerbotton, Sweden (19) (first seizure)
National Child Development Study, U.K. (16)
General Practice Research Database, U.K. (14)
Modena, Italy (9)
Copparo, Italy (10)
Lübeck, Germany (2) (first seizure)
The incidence rate in the first year of life was the highest, as in most other studies. In the second Rochester study (8), the incidence rate in the first year of life was lower than in the following 4 years. This is most probably due to the retrospective design and was not shown in the first Rochester study (4,5). Retrospective studies show a tendency to record incidence rates <100/100,000 for the first year of life (4,5,8), most probably due to incomplete case ascertainment. Although different studies reported highly variant incidence rates for the first year of life, most studies recorded rates between 110 and 170/100,000, which is well within the range of our 95% confidence interval. The two studies, one from Italy (10) and one from rural Iceland (21), recorded very high rates >200/100,000. This might be due to the higher rate of obstetric complications in their samples, as perinatal brain injuries were the most frequent etiology in the Copparo study (10), and strokes and cerebral palsy were a relatively common etiology in Iceland (21).
Incidence rates in boys and girls were similar in our study. Most other studies that assessed differences between girls and boys did not show any significant difference (4,5,10,21,23), except the study from Copparo (10), which did not explain this difference.
Epilepsies and epileptic syndromes
Concerning the incidence of epilepsies and epileptic syndromes classified along the diagnostic criteria of the ILAE (25), only a few studies are comparable (8,20,34–36). Prevalence studies on epilepsies and epileptic syndromes (37,38) are not to be compared, because they tend to underreport the incidence of those epilepsies in childhood that have a benign course. One study in France (18) classified the epilepsies after the first seizure, which leads to an ∼33% overestimation of incidence (5).
Focal versus generalized epilepsies
Similar percentages (58%) for focal epilepsies in newly diagnosed epilepsies were recorded by two studies (8,34) from the United States (55%, 58%) and by one study (20) from Sweden (52%). The slightly higher rate in the New York study (36) is due to exclusion of children with childhood absence epilepsy in this study. The lower percentage in the Dutch cohort (35) may be due to the two-stage design of this study, including children with a second seizure during a longer-lasting follow-up. It may be concluded that in western countries, focal epilepsies represent ∼55% of the total annual incidence cases in children.
Generalized epilepsies were diagnosed in 39% of the children in our study, compared with only 12–29% of the cases in the U.S. studies (8,34,36). This lower percentage is due to the high percentage of undetermined epilepsies (12–26%) in the U.S. studies besides the previously mentioned exclusion of childhood absence epilepsies in the New York study (36). In our study, we found undetermined epilepsies in only 3%; similarly, the Dutch study (35) found only 1%. The higher percentage of undetermined epilepsies in the U.S. studies might be due to incomplete data for an exact diagnosis at the time of the second unprovoked seizure.
Idiopathic versus symptomatic/cryptogenic epilepsies, specific syndromes
In our study, we recorded idiopathic epilepsies in 47% of the children. This is comparable to that reported in the Dutch study (35; 49%). However, of the idiopathic epilepsies in the Dutch study, 14% were diagnosed as focal, 86% as generalized, whereas in our study, 47% were diagnosed as focal, and 53% as generalized. This is most probably due to the low percentage of childhood absence epilepsy in our study, which was diagnosed in only 3% of all cases, whereas in the Dutch study, it was 10%, and in the Connecticut study (34), 12% of the children were diagnosed with childhood absence epilepsy. Our incidence rate for childhood and juvenile absence epilepsy (5.1/100,000) is lower than that reported in a study from Sweden focusing on absence epilepsies (39), where a rate of 6.3/100,000 in children aged 0–15 years was recorded. Childhood absence epilepsies in our study might be underreported by chance, because the reference population was relatively small, or because parents might not recognize absences as epileptic seizures and therefore might not consult a medical doctor.
The overall percentage of idiopathic epilepsies (70%) in a study from Sweden (20) was higher than that in our study (47%), whereas in the U.S. studies (8,34,36), the percentage was far lower, ranging between 20 and 30%. This might partly be due to the exclusion of childhood absence epilepsy. However, in all U.S. studies, idiopathic generalized epilepsies were less often diagnosed than in the European studies, whereas the idiopathic focal epilepsies were diagnosed in similar percentages around 10%(8,20,34–36). Some idiopathic generalized epilepsies might be classified as undetermined epilepsies in the United States, whereas in Europe, the concept of “other idiopathic generalized epilepsy” seems to be more common. In our study, 14% of the children, and in the Dutch study (35), 22% were diagnosed with “other idiopathic generalized epilepsy,” whereas in the studies from Connecticut (34) and New York (36), only 3–4% of the children received this diagnosis. In our study, all children with “other generalized idiopathic epilepsy” were classified as an “early infantile epilepsy with generalized tonic–clonic seizures.” It remains to be discussed whether this is an epileptic syndrome that should be included in the ILAE classification of epilepsies and epileptic syndromes (31,32).
Concerning the specific epilepsies and epileptic syndromes, benign childhood epilepsy with centrotemporal spikes was diagnosed in 33% of the focal epilepsies in our study, almost twice as often as in the United States (34,36), with 16%, and more often than in Sweden (20), with 21%. In our study, 22% of the total cases of newly diagnosed children were classified as having benign childhood epilepsy with centrotemporal spikes. These different figures reflect a preference to diagnose benign childhood epilepsy with centrotemporal spikes over cryptogenic focal epilepsies when a first diagnosis is made (40,41). A follow-up of the cases will clarify whether the course of the epilepsies will be benign, thus confirming the current diagnoses.
West syndrome and Lennox–Gastaut syndrome were not diagnosed in our study, but were found by others in 1–4% of cases (34–36). With 11% of all children and an incidence of 7/100,000, epilepsies with myoclonic astatic seizures were much more common in our study than in the other studies (1,34–36), with only 2% of all cases. This might be due to the difficulties in the classification and differentiation of Lennox–Gastaut syndrome and epilepsy with myoclonic astatic seizures (42). However, the high proportion of epilepsies in first-degree relatives in our children with myoclonic astatic epilepsy (75%) goes with the correct classification of our cases as epilepsies with myoclonic astatic seizures, as a strong genetic tendency is found in epilepsy with myoclonic astatic seizures, but not in Lennox–Gastaut syndrome (42,43).
The results of studies from western countries, including our study, on the total annual incidence of epilepsies and epileptic syndromes in children aged 1 month to <15 years, are similar at ∼60/100,000. Equally, they report the highest incidence rates in the first year of life and show no difference of the incidence in boys and girls. The percentage of focal epilepsies is similar, accounting for ∼55% of the total in all studies, whereas the percentages of generalized and undetermined epilepsies differs between the United States and Europe. Specific results of our study are the frequently diagnosed strongly genetically determined epilepsies with myoclonic astatic seizures, and the frequent diagnosis of early infantile epilepsy with generalized tonic–clonic seizures, in the ILAE classification subsumed under “other idiopathic generalized epilepsy.”
Acknowledgment: This work was supported by the “Verein zur Erforschung der Epidemiologie der Epilepsien,” Hamburg, Germany. We thank the private pediatricians in Heidelberg and Mannheim for their cooperation.