Incidence of childhood and youth epilepsy: A population‐based prospective cohort study utilizing current International League Against Epilepsy classifications for seizures, syndromes, and etiologies

Abstract

Objective

This study was undertaken to describe incidence and distribution of seizures, etiologies, and epilepsy syndromes in the general child and youth population, using the current International League Against Epilepsy (ILAE) classifications.

Methods

The study platform is the Norwegian Mother, Father, and Child Cohort Study (MoBa). Epilepsy cases were identified through registry linkages facilitated by Norway's universal health care system and mandatory reporting to the Norwegian Patient Registry. A standardized protocol guided medical record review, leading to validation of diagnoses and classification of seizures, epilepsy types, syndromes, and etiologies based on the latest ILAE criteria.

Results

MoBa included 111 365 participants aged 12–21 years by the end of follow‐up on December 31, 2020. We identified 1053 children and youth with epilepsy (CYE). A defined epilepsy syndrome and/or identified etiology was found in 76% of CYE in this population‐based study. Seizure types exhibited variation by age at onset. Focal epilepsies were predominant, occurring in 61% of CYE, whereas generalized epilepsies were identified in 24% of CYE. Standard clinical assessment identified etiology in 30% of CYE and in 55% with onset age < 2 years. Structural and identified genetic etiologies constituted 21% and 10%, respectively. Including presumed genetic and rare etiologies, 53% exhibited known etiology. A defined ILAE epilepsy syndrome was found in 53% of CYE. The cumulative incidence per 1000 children of the following ILAE epilepsy syndrome groups were as follows: self‐limited epilepsies, 2.25; idiopathic generalized epilepsies, 1.75; and developmental and/or epileptic encephalopathies, 2.62.

Significance

Using the new ILAE classifications, this population‐based childhood study provides incidences of seizures, epilepsies, and epilepsy syndromes. Half of epilepsy cases are classified as an ILAE epilepsy syndrome with its prognostic and therapeutic implications, but a substantial proportion of cases still have unknown etiology.

Key points.

• Seizure distribution varies by age at onset. Multiple seizure types are common in individuals with early onset epilepsy.

• Focal epilepsies predominate at all ages; combined generalized and focal epilepsies are seen mainly with early onset, whereas generalized epilepsy rises in youth.

• Identified etiology is found in 30% of CYE, most commonly structural etiology in 21% and identified genetic etiology in 10%.

• When including presumed genetic etiology, 53% of epilepsies are classified as known etiology.

• An ILAE epilepsy syndrome is identified in 53% of CYE, and three quarters of CYE have a known etiology and/or an epilepsy syndrome.

1. INTRODUCTION

Epilepsy, defined by recurrent unprovoked epileptic seizures, imposes the greatest disease burden among chronic neurological conditions in children. ¹ , ² The International League Against Epilepsy (ILAE) has introduced a classification system spanning three diagnostic levels: seizure type, epilepsy type, and the newly added epilepsy syndromes. ³ This novel classification underscores the importance of considering etiology at each diagnostic level due to its profound implications for treatment. ³

Seizure classification is based on practical criteria, categorizing seizures by onset feature into focal, generalized, or unknown onset. ⁴ A distinguishing characteristic of focal seizure is awareness. ⁴

Following seizure classification, epilepsy types are determined, including focal epilepsy, generalized epilepsy, combined generalized and focal epilepsy, and an unknown epilepsy category indicating the presence of epilepsy with an unidentified seizure onset. ³

Epilepsy syndrome classification relies on distinct clinical and electroencephalographic (EEG) features, often supplemented by etiological findings. ⁵ Syndromes are classified based on seizure types within different age at onset categories. ³ A separate category, developmental and/or epileptic encephalopathy (DEE), denotes an epilepsy associated with developmental impairment attributable to either the underlying etiology, the superimposed epileptic activity, or both. ³

Etiological categories encompass structural, genetic, infectious, metabolic, immune, and unknown causes, permitting possible combinations. ⁵

Our current understanding of the prevalence and distribution of seizures, syndromes, and etiologies in the broader pediatric population remains limited. ⁶ , ⁷ , ⁸ , ⁹ A comprehensive international meta‐analysis of epilepsy incidence and prevalence reveals significant heterogeneity between studies, with few studies focusing on seizure type or etiology. ¹⁰ This underscores the need for research based on the Standards for Epidemiologic Studies and Surveillance of Epilepsy. ¹⁰ The Nova Scotia, Minnesota, Connecticut, and Finnish studies are important population‐based studies that have provided comprehensive insight into childhood onset epilepsy, but not fully incorporating the current classifications. ¹¹ , ¹² , ¹³ , ¹⁴ The Dutch study of epilepsy in childhood, while providing longitudinal long‐term data on seizure outcomes, is hospital‐based and also lacks detail regarding the latest classification. ¹⁵ The formalization of ILAE criteria for epilepsy syndromes represents a notable update to previous classifications, aiming to improve and harmonize our understanding of epilepsy syndromes. ⁵ Recent population‐based studies, while incorporating these classification updates and diagnostic advancements, primarily focus on early onset epilepsies. ⁶ , ¹⁶ , ¹⁷ This gap in the knowledge about the onset and progression of childhood and youth epilepsies, particularly regarding advancements in diagnostics and updated classifications, needs to be addressed. To contribute to this, we present population‐based incidence data on epileptic seizures, etiologies, and epilepsy syndromes in children and youth, using the Norwegian Pediatric Epilepsy Cohort (NorPEC).

2. MATERIALS AND METHODS

2.1. Study population

The study population is based on the Norwegian Mother, Father and Child Cohort Study (MoBa), a nationwide, prospective, population‐based pregnancy cohort with recruitment between 1999 and 2008. ¹⁸ , ¹⁹ The mothers consented to participation in 41% of pregnancies, recruiting approximately 114 500 children before birth and prior to epilepsy diagnoses. ¹⁸ , ¹⁹ Through linkage to the Medical Birth Registry, a national health registry containing information about all births in Norway, we selected eligible participants as all live‐born MoBa children residing in Norway until either their demise or the conclusion of registry follow‐up on December 31, 2020.

2.2. Epilepsy patient ascertainment

Within MoBa, we established a case–cohort study of epilepsy: NorPEC. Potential epilepsy cases were identified by linking MoBa to the Norwegian Patient Registry (NPR) using International Classification of Diseases, 10th Edition (ICD‐10) codes G40.X/G41.X. ²⁰ The NPR collects mandatory data from all secondary and tertiary health care services in Norway, covering both public and private hospitals and practitioners, integrated with government reimbursement. ²¹ It is common practice to refer all children with certain or suspected epilepsy to these services for clinical evaluation and EEG, as stated in national guidelines. ²² Since 2008, the NPR has been made personally identifiable and searchable, facilitating robust analysis and research in health care. ²¹ Children registered with at least one recorded ICD‐10 G40/41 code in the NPR by December 31, 2020 were included as potential epilepsy cases. In an earlier data collection, potential epilepsy cases prior to the NPR's transition to personally identifiable data were identified by MoBa questionnaire report of epilepsy, and these were carried forward to the current study. ⁶ Medical records were reviewed using a modified standardized protocol, which included inquiries about medical and developmental history, additional diagnoses, age at seizure onset, seizure descriptions, frequency, investigation outcomes, treatment, and treatment responses. ⁶ , ²³ Children and youth with epilepsy (CYE) were defined by the operational clinical ILAE definition of epilepsy, comprising participants with: “1; At least two unprovoked (or reflex) seizures occurring more than 24 h apart, 2; One unprovoked (or reflex) seizure and a likelihood of further seizures similar to the general recurrence risk (at least 60%) after two unprovoked seizures, occurring over the next 10 years, or 3; Diagnosis of an epilepsy syndrome.” ³

Epilepsy onset age was defined as the age at which a patient experienced their first unprovoked epileptic seizure and subsequently met the criteria for an epilepsy diagnosis. Epileptic seizures, epilepsy types, and syndromes were classified according to the new ILAE classifications, utilizing all available clinical and investigation results throughout the study period. We sought a diagnosis at all three levels, as well as the etiology, where possible. We classified as DEE cases where global developmental delay was evident in CYE or an ILAE syndrome is formally classified in the DEE category. Our approach reflects the standard diagnostic process conducted for each patient in the clinical setting at the time of investigation. This entails evaluations carried out at the hospital during both the initial diagnosis and subsequent follow‐up visits throughout the study period. Classification was conducted by two pediatric epileptologists (T.V. and K.M.A.), and any disparities in opinion were resolved through consensus.

In reporting the study, we adhered to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines. ²⁴

2.3. Statistical methods

We analyzed the data using Stata/SE 17.0 for Windows, updated May 16, 2023. Instances where the outcome was not observed before the end of follow‐up were treated as censored observations (e.g., a child followed up to age 13 years is censored for outcomes from age 13 onward), except for percentages of all CYE. Age at epilepsy onset was missing for two CYE, and these are excluded from age‐dependent analyses. We categorized cases by seizures, epilepsy types, etiologies, and epilepsy syndromes. Incidence rates were calculated by age, with associated 95% confidence intervals (CIs). The Nelson–Aalen estimator calculates cumulative incidence, excluding the 347 nonepilepsy deaths (only available dichotomized; .2% before and .1% after 12 months of age). Cumulative incidence per 100 000 children 0–18 years old is reported where not otherwise specified. To explore differences across categorical variables, Pearson chi‐squared tests were employed with two‐sided p‐values (Fisher exact test where applicable). Given underrepresentation of certain sociodemographic groups in MoBa, ¹⁸ , ¹⁹ , ²⁵ we examined their impact on epilepsy incidence. Using Poisson regression, we assessed the influence of parental age, education, marital status, pregnancy planning, parity, and mother's country of birth. Model fit and robustness were evaluated through overdispersion checks and stability analyses.

2.4. Ethics

MoBa is conducted by the Norwegian Institute of Public Health, initially based on a license from the Norwegian Data Protection Agency and approval from the Regional Committees for Medical and Health Research Ethics, and is currently regulated by the Norwegian Health Registry Act. ¹⁸ , ¹⁹ Participation requires informed consent, allowing linkages to health registries and medical record reviews. The NorPEC study holds a separate approval from the Regional Committee for Medical and Health Research Ethics (ref. 20478). To protect confidentiality, characteristics in fewer than five participants are not detailed.

3. RESULTS

The MoBa study population comprised 111 365 live‐born children, with a median age of 15.2 years (range = 11.4–21.3 years for those still alive) at registry linkage on December 31, 2020 (Figure 1). By the end of registry follow‐up on the same date, 1603 participants had an epilepsy diagnosis registered in the NPR. We confirmed epilepsy in 1053 of these participants (positive predictive value = 66%). For the 33 CYE who died during follow‐up, median age at death was 10.0 years (range = .1–20 years), and cause of death was related to the underlying etiology in the majority (28 CYE), with the remaining five distributed among seizure‐related causes, including sudden unexpected death in epilepsy, and unknown cause of death. Median age at data collection was 16.7 years (range = 1 month–22.6 years). Median age at onset of epilepsy was 6.0 years (range = 1 day–18.0 years) and was unknown in two CYE. Median follow‐up time from epilepsy onset was 10.6 years (range = 0–21 years); all but six CYE had at least 1 year of follow‐up time. EEG examinations were performed in all but four CYE, and 483 (46%) had undergone long‐term EEG recordings (336 with video‐EEGs). Cerebral magnetic resonance imaging (MRI) had been conducted in 931 (88%). Genetic testing was reported for 221 (21%) and metabolic testing for 131 (12%).

FIGURE 1.

Flowchart illustrating the recruitment process, data collection, and validation of epilepsy diagnoses. ^aIneligible participants include children not at risk for epilepsy, such as stillborn children, children who emigrated from Norway without an epilepsy diagnosis, and children with invalid personal identification numbers. Three hundred eighty‐three participants who died are also excluded, as age at death is not available for incidence calculations. ^bThe Norwegian Mother, Father and Child Cohort Study (MoBa) reports of epilepsy were included in a previous data collection, to capture cases before Norwegian Patient Registry (NPR) was made personally identifiable. ILAE, International League Against Epilepsy.

The general characteristics of the CYE are described in Table 1. There were considerable differences across age at onset categories. CYE with onset before 2 years were more likely to have a history of neonatal seizures (relative risk [RR] = 5.7, 95% CI = 3.6–8.8), febrile seizures (RR = 2.0, 95% CI = 1.5–2.8), ≥3 seizure types during follow‐up (RR = 3.7, 95% CI = 2.9–4.7), abnormal MRI (RR = 1.9, 95% CI = 1.6–2.2), and abnormal neurological examination (RR = 3.0, 95% CI = 2.5–3.5). Among the 74 CYE with a history of neonatal seizures, most experienced acute symptomatic seizures during this period, and the majority developed epilepsy afterward; only 15 had their epilepsy onset during the neonatal period. There were no apparent differences across ages at onset with regard to gender distributions (p = .76) or heredity of epilepsy (p = .19).

TABLE 1.

General characteristics of CYE.

General characteristics of CYE	CYE by age at onset of epilepsy a
	All CYE, N = 1053		0–2 years, n = 227		2–5 years, n = 217		5–12 years, n = 465		>12 years, n = 142		p b
	n	%	n	%	n	%	n	%	n	%
Male	562	53.4	123	54.2	115	53.0	252	54.2	70	49.3	.76
Family history of epilepsy c	241	22.9	46	20.3	60	27.6	99	21.3	36	25.4	.19
History of neonatal seizures	74	7.0	45	19.8	13	6.0	14	3.0	2	1.4	<.001
History of febrile seizures	138	13.1	49	21.6	44	20.3	36	7.7	9	6.3	<.001
Seizure precipitants reported d	393	37.3	113	49.8	99	45.6	123	26.5	57	40.1	<.001
1 seizure type reported	521	49.5	64	28.2	95	43.8	285	61.3	77	54.2	<.001 e
2 seizure types reported	327	31.1	61	26.9	81	37.3	135	29.0	50	35.2
≥3 seizure types reported	201	19.1	101	44.5	40	18.4	45	9.7	15	10.6
Diagnosed with epilepsy or started therapy after first unprovoked seizure f	90	8.5	8	3.5	22	10.1	44	9.5	16	11.3	.02
Only 1 seizure during follow‐up f	38	3.6	1	.4	2	.9	22	4.7	13	9.2	<.001
Abnormal MRI findings	335	31.8	113	49.8	73	33.6	114	24.5	34	23.9	<.001
Abnormal neurological exam	317	30.1	143	63.0	72	33.2	76	16.3	24	16.9	<.001
Abnormal EEG findings	934	88.7	190	83.7	181	83.4	438	94.2	123	86.6	<.001
Epileptiform EEG abnormalities	905	85.9	182	80.2	175	80.6	430	92.5	116	81.7	<.001

Abbreviations: CYE, children and youth with epilepsy; EEG, electroencephalographic; MRI, magnetic resonance imaging.

^a Age at onset of epilepsy is missing for 2 CYE.

^b Probability values calculated with Pearson chi‐squared tests (Fisher exact test where applicable) across age at onset groups.

^c Reports of epilepsy among first‐ and second‐degree relatives.

^d Sleep deprivation, flashing lights, mental or physical stress, hyperventilation, fever, sudden noise.

^e Probability value calculated with Pearson chi‐squared test across age at onset groups and number of seizures categories.

^f Only seizures that are considered confirmed based on clinical judgment during medical record review are counted.

3.1. Nonepilepsy cases after validation

Of the children and youth registered with at least one epilepsy diagnosis in the NPR, 550 did not meet the operational clinical definition of epilepsy. Reasons for evaluation included seizurelike events and nonseizurelike challenges such as concentration difficulties or language problems. Of confirmed nonepilepsy cases, 11% reported epileptic seizures (provoked or single seizures without at least 60% recurrence risk or a defined epilepsy syndrome), whereas 68% experienced other seizurelike events such as unspecific blank spells (25%), febrile seizures (9%), or syncope (8%). Notably, only 13 confirmed noncases had functional/dissociative seizures. The source of NPR epilepsy diagnoses in confirmed nonepilepsy cases varied, including clinician diagnosis (20%), provisional clinician diagnosis later discredited (20%), with the remaining diagnoses distributed among EEG laboratory, administratively registered diagnoses, miscoding, and unknown source.

3.2. Incidence rate and cumulative incidence of epilepsy

Table 2 presents epilepsy incidence rates, with detailed age‐group data in Table S1. Figures 2 and 3 display the incidence rate and cumulative incidence of epilepsy. The incidence rate was highest in the first year of life at 149 per 100 000 person‐years (95% CI = 127–173). Cumulative incidence per 100 000 children was 226 (95% CI = 199–508) at 2 years, 414 (95% CI = 378–454) at 5 years, 827 (95% CI = 775–883) at 12 years, and 1047 (95% CI = 980–1118) at 18 years. A tendency toward higher incidence in boys was not statistically significant (odds ratio = 1.09, 95% CI = .96–1.23). Incidence rate by categories of epilepsy syndromes exhibited age‐dependent patterns, with DEE epilepsy syndromes accounting for three fourths of incident cases in the first year of life. Univariate models showed a tendency for both maternal and paternal education levels to influence epilepsy incidence, but in the multivariate model including other sociodemographic variables, only paternal education level significantly influenced epilepsy incidence, with >12 years of education associated with a lower incidence rate (Poisson coefficient = −.206, incidence rate ratio = .814, 96% CI = .648–.923, p = .007).

TABLE 2.

Distribution of epilepsy types, epilepsy syndromes, and etiologies.

Epilepsy Type/Syndrome/Etiology	n	% of CYE	Incidence rate per 100000 person years (95% CI)
All epilepsies	1053	100	61.8 (58.2–65.7)
Focal epilepsy	639	61	37.5 (34.7–40.5)
Generalized epilepsy	254	24	14.9 (13.2–16.9)
Combined generalized and focal epilepsy	82	7.8	4.8 (3.9–6.0)
Unknown epilepsy type	78	7.4	4.6 (3.7–5.7)
Epilepsy syndromes	553	52	32.5 (29.9–35.3)
GGE	221	21	13.0 (11.4–14.8)
IGE	154	15	9.0 (7.7–10.6)
CAE	83	7.9	4.9 (3.9–6.0)
JAE	31	2.9	1.8 (1.3–2.6)
JME	24	2.3	1.4 (.9–2.1)
GTCA	22	2.1	1.3 (.9–2.0)
Self‐limited epilepsy syndromes a	245	23	14.4 (12.7‐16.3)
Self‐limited neonatal and infantile epilepsies	26	2.5	1.5 (1.0–2.2)
GEFS+	26	2.5	1.5 (1.0–2.2)
SeLECTS	163	15	9.6 (8.2–11.2)
SeLEAS	37	3.5	2.2 (1.6–3.0)
COVE	7	.66	.4 (.2–.9)
DEE b	269	26	15.8 (14‐17.8)
Defined ILAE DEE syndromes	116	11	6.8 (5.7–8.2)
EIDEE	9	.85	.5 (.3–1.0)
IESS	48	4.6	2.8 (2.1–3.7)
Dravet syndrome	6	.57	.4 (.2–.8)
Lennox–Gastaut syndrome	16	1.5	.9 (.6–1.5)
DEE‐SWAS	13	1.2	.8 (.4–1.3)
Etiology‐specific syndromes c	8	.76	.5 (.2–.9)
Etiologies
Known	555	53	32.6 (30.0–35.4)
Identified	312	30	18.3 (16.4–20.5)
Structural	224	21	13.1 (11.5–15)
Hypoxic–ischemic encephalopathy	54	5.1	3.2 (2.4–4.1)
Vascular sequelae	44	4.1	2.6 (1.9–3.5)
Malformation of cortical development	39	3.7	2.3 (1.7–3.1)
Intracerebral tumor	26	2.5	1.5 (1.2–2.0)
Infection sequelae	13	1.2	.8 (.4–1.3)
Trauma sequelae	8	.76	.5 (.2–.9)
Vascular malformation	8	.76	.5 (.2–.9)
Mesial temporal sclerosis	5	.47	.3 (.1–.7)
Other	26	2.5	1.5 (1.0–2.2)
Genetic	351	33	20.6 (18.6–22.9)
Identified genetic	101	9.6	5.9 (4.9–7.2)
Single gene variant	63	6.0	3.7 (2.9–4.7)
Copy number variation	25	2.3	1.5 (1.0–2.2)
Chromosomal	6	.57	.4 (.2–.8)
Presumed genetic	248	24	14.6 (12.9–16.5)
Infectious	15	1.4	.9 (.5–1.5)
Metabolic	7	.66	.4 (.2–.9)
Immune	2	.19	.1 (0–.5)
Unknown	498	47	29.2 (26.8–31.9)
Focal structural epilepsy, all	226	21	13.3 (11.6–15.1)
Focal structural epilepsy, no DEE	118	11	6.9 (5.8–8.3)

Abbreviations: CAE, childhood absence epilepsy; CI, confidence interval; COVE, childhood occipital visual epilepsy; CYE, children and youth with epilepsy; DEE, developmental and/or epileptic encephalopathy; DEE‐SWAS, DEE with spike–wave activation in sleep; EIDEE, early infantile epileptic encephalopathy; GEFS+, genetic epilepsy with febrile seizures plus; GGE, genetic generalized epilepsy; GTCA, generalized epilepsy with tonic–clonic seizures alone; IESS, infantile epileptic spasms syndrome; IGE, idiopathic generalized epilepsy; ILAE, International League Against Epilepsy; JAE, juvenile absence epilepsy; JME, juvenile myoclonic epilepsy; SeLEAS, self‐limited epilepsy with autonomic seizures; SeLECTS, self‐limited epilepsy with centrotemporal spikes.

^a Only specific ILAE syndromes.

^b All DEEs, including those without defined ILAE DEE syndromes.

^c Includes neonatal onset syndromes (KCNQ2‐DEE, PD‐DEE/PNPO‐DEE, CDKL5‐DEE, PCDH19 clustering epilepsy, Glut‐1 deficiency syndrome, Sturge–Weber syndrome, and gelastic seizures with hypothalamic hamartoma).

FIGURE 2.

Incidence rate per 100 000 person‐years. The shaded area represents the associated 95% confidence interval.

FIGURE 3.

Cumulative incidence of epilepsy and selected categories. International League Against Epilepsy (ILAE) syndrome categories: genetic generalized epilepsy (GGE), self‐limited epilepsy syndromes (include neonatal/infantile onset self‐limited epilepsies and childhood self‐limited focal epilepsies). The categories developmental and/or epileptic encephalopathy (DEE) and focal structural epilepsy encompass epilepsies that are and are not classified as defined ILAE syndromes.

3.3. Epileptic seizures

Table 3 presents the distribution of all seizures experienced by CYE during follow‐up, with age at epilepsy onset details in Table S2. Because 50.5% of patients had multiple seizure types, column totals exceed 1053. Seizure types per CYE ranged from zero to seven, with a median of two. Two CYE had no verified seizures, both among the 13 with DEE with spike–wave activation in sleep (DEE‐SWAS), where seizures are not mandatory for an epilepsy diagnosis. ²⁶

TABLE 3.

Distribution of epileptic seizure types.

Seizure type	Population proportion per 100 000 children a	All CYE b
		1053
		n	%
Focal onset seizures	649	725	69
Focal onset aware	136	152	14
Focal onset impaired awareness	451	504	48
Focal to bilateral tonic clonic	363	406	39
Motor onset	354	395	38
Automatisms	55	62	5.9
Atonic	21	24	2.3
Clonic	164	183	17
Spasms, focal onset	17	19	1.8
Hyperkinetic	7	8	0.8
Myoclonic	9	10	0.9
Tonic	88	98	9.3
Nonmotor onset	238	266	25
Autonomic	83	93	8.8
Behavioral arrest	44	49	4.7
Cognitive	19	21	2
Emotional	7	8	0.8
Sensory	76	85	8.1
Generalized onset seizures	303	339	32
Motor	198	221	21
Tonic‐clonic	157	175	17
Clonic	0	–	0
Tonic	59	66	6.3
Myoclonic	91	102	10
Myoclonic tonic‐clonic	4	4	0.4
Myoclonic atonic	2	2	0.2
Atonic	27	30	2.8
Spasms, generalized onset	16	18	1.7
Nonmotor (absences)	192	215	20
Typical absences	134	150	14
Atypical absences	21	23	2.2
Other absences (myoclonic, eyelid myoclonia, other)	39	44	4.2
Unknown onset seizures	100	112	11
Motor	79	88	8.4
Tonic‐clonic unknown onset	43	48	4.6
Epileptic spasms, unknown onset	16	18	1.7
Nonmotor (behavioral arrest)	27	30	2.8
Across all seizure onset categories
All spasms	49	55	5.2
All tonic‐clonic	557	622	59
Both generalized and focal	74	83	7.9

Abbreviation: CYE, children and youth with epilepsy.

^a Population N = 111 365.

^b Column totals add up to >1053 due to multiple seizure types possible in an individual. For age‐specific distribution, see Table S2.

Focal onset seizures were the most frequent seizure type in all epilepsy onset age groups (p < .001), reported in 69% of all CYE, 77% with onset before age 2 years, falling to 51% beyond age 12 years. The majority of focal onset seizures were impaired awareness seizures. Motor onset focal seizures were more common than nonmotor onset. Focal to bilateral tonic–clonic seizures were reported in 39% of CYE, with no significant differences in occurrence across age at onset categories (p = .70).

During follow‐up, 32% of CYE had generalized onset seizures, with tonic–clonic seizures in 17%, typical absences in 14%, myoclonic seizures in 10%, and tonic seizures in 6.3%. The distribution of generalized onset seizures showed a bimodal pattern, being more common in individuals with epilepsy onset before the age of 2 years (38%) and after age 12 years (42%). Typical absences were more frequent in older onset groups, particularly ages 5–12 years, accounting for 20% of CYE in this group. Some seizure types correlated with younger epilepsy onset: tonic seizures in 20% of children with onset before age 2 years, myoclonic in 23%, and atonic in 8%.

CYE with neonatal/infantile epilepsy onset had a higher proportion of unknown mode of onset seizures (p < .001), most commonly epileptic spasms (p < .001) and other motor onset seizures (p < .001).

Overall, 622 CYE (59%) had a history of tonic–clonic seizures during follow‐up of epilepsy, including focal, generalized, and unknown onset types. Tonic–clonic seizures were most frequent in CYE with DEE (187/266, 70%), focal structural epilepsies (146/223, 65%), and self‐limited epilepsy with centrotemporal spikes (103/163, 63%). Epileptic spasms were evenly distributed among focal, generalized, and unknown onset seizures, with a total of 55 (5.2%) CYE experiencing spasms, significantly correlated with younger epilepsy onset (23% of CYE in the neonatal/infantile onset group). Forty‐nine of 55 CYE with epileptic spasms had epilepsy onset within the first year of life and the remaining six before age 3 years. A total of 83 CYE (7.9%) experienced both focal and generalized seizures, significantly more frequent among younger epilepsy onset (24% of CYE with neonatal/infantile onset). Status epilepticus was reported in 190 (18% of CYE).

Ninety CYE (8.5%) received their diagnosis or started antiseizure medication after their first unprovoked seizure; of these, 52 (58%) experienced subsequent seizures. Thirty‐eight CYE (3.6%) had only one certain seizure during follow‐up, most commonly in self‐limited focal epilepsies (17 CYE) and structural etiologies (11 CYE).

3.4. Etiology

As shown in Table 2, the major etiological categories were structural and genetic etiologies. Structural etiology, where abnormal structural cerebral MRI findings and electroclinical assessment lead to a reasonable inference that the imaging abnormality is the likely cause of the patient's seizures, occurred in 21% of CYE, the most common of which were sequelae after hypoxic–ischemic encephalopathy (5.1% of CYE), sequelae after cerebral hemorrhage or infarction (vascular sequelae; 4.2% of CYE), and malformations of cortical development (3.7% of CYE), followed by cerebral tumors (2.5% of CYE). In only .47% of CYE, mesial temporal sclerosis was identified on first or later cerebral MRI during follow‐up (one identified on histology only). We have divided genetic etiology into molecularly identified genetic etiology in 10% of CYE and presumed genetic etiology (24% of CYE) based on a family history suggestive of autosomal dominant inheritance or a specific epilepsy syndrome where population studies have suggested a genetic basis. Identified genetic etiology was most frequently variants in SCN1A and trisomy 21 (both nine CYE). Sixty‐three CYE had single gene disorders, and variants in the following genes were found in more than one CYE: SCN1A, GRIN2A, MECP2, CDKL5, STXBP1, PCDH19, and CACNA1A. Identified infectious, metabolic, and immunologic etiologies were rare (24 CYE). Thirty‐eight CYE (3.6%) belonged to more than one etiological category, most frequently structural and genetic (20 CYE, six with tuberous sclerosis complex) or structural and infectious etiology (13 CYE, six due to cytomegalovirus). The proportion of CYE classified as having an identified etiology (structural, identified genetic, infectious, metabolic, and/or immunological) was 30% overall, but when restricting to those with onset age < 2 years, this proportion was 53%. When including those with presumed genetic etiology, the proportion classified as known etiology was 53% of all CYE. Correspondingly, etiology was classified as unknown in 47%.

3.5. Epilepsies and syndromes

The distribution of epilepsy types and syndromes according to the ILAE epilepsy classification ⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ is presented in Table 2. We first classified epilepsy type according to the mode of seizure onset as focal in 639 (61%), generalized in 254 (24%), combined generalized and focal in 82 (7.8%), and unknown in 78 (7.4%).

At the epilepsy syndrome level, 553 (53%) were classified as a defined ILAE epilepsy syndrome. The most frequent specific epilepsy syndromes were self‐limited epilepsy with centrotemporal spikes (SeLECTS) in 163 CYE (cumulative incidence = 148 per 100 000 children), followed by childhood absence epilepsy (CAE) in 83 CYE (cumulative incidence = 75).

In the broader categories of ILAE epilepsy syndromes, self‐limited epilepsy syndromes were seen in 245 (23% of CYE, cumulative incidence = 225).

The cumulative incidence of generalized genetic epilepsy syndromes (GGE) was 238, of which 175 were idiopathic generalized epilepsies (IGEs). For the specific IGE syndromes, cumulative incidences were 75 for CAE, 34 for juvenile absence epilepsy (JAE), 36 for juvenile myoclonic epilepsy (JME), and 20 for generalized epilepsy with tonic–clonic seizures alone (GTCA). Six patients evolved from CAE to another IGE with increasing age. The remaining GGEs were epilepsy with eyelid myoclonia, epilepsy with myoclonic absences, epilepsy with myoclonic–atonic seizures, and GGE not categorized into specific ILAE epilepsy syndromes.

DEEs were seen in 269 (26% of CYE), 116 (11.0% of CYE, 43% of DEEs) of whom had a specific ILAE DEE epilepsy syndrome. DEE was more common in early onset epilepsies, with 120 cases (72% of CYE) in the 0–1 year age group and 40 cases (47% of CYE) in the 1–2 year age group. Forty‐four percent of epilepsies with onset in the first year of life were specific ILAE DEE epilepsy syndromes; the most frequent across all ages were 48 CYE with infantile epileptic spasms syndrome (IESS), 16 with Lennox–Gastaut syndrome (12 of which evolved from IESS), 13 with DEE‐SWAS, and six with Dravet syndrome. The etiology‐specific DEE syndromes are rare but together constitute eight CYE.

4. DISCUSSION

In this unique longitudinal, population‐based, national study, we provide contemporary data on children and youth aged 0–18 years on the incidence and distribution of seizures, epilepsies, etiologies, and epilepsy syndromes according to the latest ILAE classifications. The study builds on previous work from our group. ⁶ , ¹⁶ , ³⁰ A defined syndrome and/or etiology is identified in 76% of the epilepsies in the current cohort.

The incidence rate of epilepsy in children in population‐based studies from high‐income countries has ranged from 33.3 to 82 cases per 100 000 persons per year. ⁹ Fiest's meta‐analysis reported a pooled incidence rate of 46.90 (95% CI = 42.29–52.01) per 100 000 person‐years 0–18 years old and a pooled annual cumulative incidence of 85.29 (95% CI = 59.54–122.19) per 100 000 persons 0–18 years old; however, there were few included studies that reported on children, making that estimate uncertain. ¹⁰ The meta‐analysis included studies from low‐ and middle‐income countries, where the incidence of epilepsy is expected to be higher due to factors such as perinatal complications, infection, and trauma. ³¹ However, universal identification of cases in these regions may be more challenging. Our finding of a higher incidence rate of 61.8 (95% CI = 58.2–65.7) may be attributed to our population‐based and longitudinal study design, universal and free health care for children, and the use of the new definition of epilepsy. Our cumulative incidence of 1046 (95% CI = 980–1118) at age 18 years contributes to estimates through childhood in a high‐income country.

4.1. Distribution of epileptic seizures

As with previous studies, focal onset seizures are most frequent at all ages of epilepsy onset. ⁷ , ¹⁶ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ Predominance of early onset epilepsies probably explains our high proportions with multiple seizure types during follow‐up (19% had ≥3 seizure types) and seizure types associated with early onset (atonic, myoclonic, and tonic seizures and epileptic spasms). Our observed population proportion of spasms of 49 per 100 000 children aligns with findings from studies on IESS. ²⁷ , ³⁷

4.2. Distribution of etiologies

In the current study, there was a known etiology, including presumed genetic, in 53% of CYE. This is lower than the 63% in those aged 0–19 years in a recent Norwegian cross‐sectional population‐based study, which was limited to active epilepsy cases. ³⁸ Structural etiology was identified in 21.3% of cases, a figure that falls between those reported in the Minnesota and Connecticut studies. It shows a similar proportion of perinatal injury and cortical dysplasia, but fewer cases of mesial temporal sclerosis, possibly explained by their longer follow‐up periods, although a true decrease has been discussed in surgical cohorts. ¹² , ³⁹ , ⁴⁰ Among children aged 0–3 years, an identifiable etiology was found in 50%, similar to the 54% reported in a Scotland‐wide study. ¹⁷ The Scottish study conducted genetic testing for all children with unknown etiology, whereas our study relied solely on results from standard clinical evaluations. This suggests that although the current clinical approach is effective, there is room for improvement. The clinical practice of limiting genetic testing to more severe epilepsies is evident in our study, as 86 of 97 identified genetic etiologies were observed in CYE with DEE. Genes associated with milder forms of epilepsy, such as PRRT2 (identified in 4% of cases in the Scottish study), went unnoticed in our cohort. Expanding gene panel testing, also including early onset epilepsies that might be self‐limited, could yield more positive results due to advancements since our cohort's initial evaluations. Similarly, extending radiological testing could uncover hidden etiologies, as improved neuroimaging techniques enhance detection of epilepsy foci and understanding of neurobiological mechanisms. ⁴¹ , ⁴²

4.3. Distribution of epilepsy types and syndromes

Our distribution of epilepsy types reveals a high (61%) proportion of focal epilepsies, but as the cohort matures we anticipate identifying children with later onset generalized epilepsies, such as JAE and JME, which would reduce the overall proportion of focal epilepsies in childhood. Epidemiological data on epilepsy types vary according to study population, geographical differences, and age. ⁹ To the best of our knowledge, this is the first population‐based study to comprehensively adopt the new ILAE epilepsy syndrome classification. ⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ The most common among syndromes with neonatal and infantile onset, followed by childhood onset and IGE syndromes, are discussed below.

In our study, generalized epilepsy with febrile seizures plus is the most frequently occurring self‐limited epilepsy syndrome with onset in neonates and infants (cumulative incidence = 23 per 100 000 children). In the Scottish prospective population‐based study, the reported incidence was 5.9/100 000 live births at 0–3 years of age. ¹⁷ IESS is the most common syndrome with onset in infants and neonates (cumulative incidence = 43), similar to previous studies and supporting a slightly higher tendency described in the other Nordic countries. ²⁷ , ³⁷

SeLECTS is the most frequent syndrome among childhood focal epilepsies, with a reported annual incidence of 5.3–21/100 000 in children aged <16 years, with the higher incidence in a hospital‐based Swedish study in the 1970s. ¹² , ⁴³ , ⁴⁴ , ⁴⁵ We found a cumulative incidence of 148 and incidence rate of 10/100 000 person‐years in children aged <16 years, which are higher than previously reported in population‐based studies. ²⁶ , ⁴³ This difference may be explained by our implementation of the new clinical definition of epilepsy not requiring ≥2 unprovoked seizures in the presence of an epilepsy syndrome. We found that 16 of 163 SeLECTS cases experienced only one observed seizure, aligning with the reported 10% presenting only one seizure in this syndrome. ⁴⁶ A population‐based design with universal health care, where even mild epilepsy is evaluated, may also contribute to the higher incidence.

The definition of IGEs varies across studies, with differences in clinical and EEG criteria making comparisons challenging, which is reflected in the 6.4%–35.8% of epilepsies reported. ⁴⁷ The current classification restricts IGEs to four syndromes (CAE, JAE, JME, and GTCA). Our cumulative incidence of CAE of 75 is similar to previous studies, suggesting that it is likely we captured most CAE cases in our cohort. ⁴⁷ Incidence rates of juvenile onset IGE in our cohort are lower than previously documented, but we expect an increase as the cohort matures.

4.4. Strengths and limitations

This study has several strengths. It is a population‐based pregnancy cohort design with longitudinal follow‐up, large sample size, and nationwide recruitment in a country with universal health care. It applies a new formally accepted classification scheme on CYE with clinically validated epilepsy. We have a long follow‐up time for childhood onset epilepsies. However, the study also has limitations, the most important one being that some of the data were collected several years after epilepsy onset. Data from the NPR were made personally identifiable in 2008, and diagnoses registered exclusively before this time might not be completely captured by MoBa questionnaires. Furthermore, the investigations are standard clinical evaluations conducted at the time of diagnosis and follow‐up, as the researchers rely solely on these existing assessments, which are not part of a strictly regulated research protocol.

Our MoBa sample has somewhat higher parental education levels and fewer immigrants, non‐Caucasians, and single mothers than the general population. ¹⁸ , ¹⁹ , ²⁵ Most sociodemographic variables were not linked to epilepsy incidence, but paternal education was, possibly underestimating epilepsy incidence in lower education groups, suggesting slightly higher incidence in the general population.

In Norway, health care is universally accessible and free of charge for children, and diagnostic evaluations for suspected epilepsy in childhood are conducted in secondary and tertiary health care, with mandatory reporting to the NPR. Therefore, most cases are believed to be captured, although underreporting of infrequent or unnoticed seizures is possible.

5. CONCLUSIONS

This study delineates incidence of epilepsy according to the new ILAE classification, including epilepsy syndromes. The age at onset significantly influenced the distribution of seizures, epilepsies, and epilepsy syndromes, as well as their respective etiological and developmental categories. Despite advancements in diagnostics, the etiology remained unknown in half of children and youth with epilepsy. Children with younger onset epilepsy were more likely to have an identifiable etiology, predominantly attributed to structural and genetic factors. Three quarters of CYE in this Norwegian cohort had an epilepsy of known etiology and/or a defined ILAE epilepsy syndrome.

AUTHOR CONTRIBUTIONS

Truls Vikin: Formal analysis (lead); funding acquisition (equal); data collection (equal); methodology (equal); project management (equal); validation of clinical data (equal); writing—original draft preparation (lead); writing—review and editing (equal). Morten I. Lossius: Funding acquisition (supporting); methodology (equal); supervision (supporting); writing—review and editing (equal). Ragnhild E. Brandlistuen: Methodology (supporting); writing—review and editing (equal). Richard F. Chin: Conceptualization (lead; equal); methodology (equal); writing—review and editing (equal). Kari M. Aaberg: Conceptualization (lead; equal); formal analysis (supporting); funding acquisition (equal); data collection (equal); methodology (equal); project management (lead); supervision (lead); validation of clinical data (equal); writing—original draft (supporting); writing—review and editing (equal).

CONFLICT OF INTEREST STATEMENT

T.V. has received a speaker fee from Eisai unrelated to the submitted work. K.M.A. has received a speaker fee from UCB unrelated to the submitted work. R.F.C. has provided paid consultancy for GW Pharma and Zogenix and has received travel grants and speaker fees from GW Pharma, Zogenix, Eisai, and UCB. M.I.L. has provided paid lectures and served on expert panels for UCB, Eisai, Jazz, and Angelini. R.E.B. has no conflicts of interest. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

CONSENT

Participation in the Norwegian Mother, Father, and Child Cohort is based on informed consent.

Supporting information

Table S1.

Vikin T, Lossius MI, Brandlistuen RE, Chin RF, Aaberg KM. Incidence of childhood and youth epilepsy: A population‐based prospective cohort study utilizing current International League Against Epilepsy classifications for seizures, syndromes, and etiologies. Epilepsia. 2025;66:776–789. 10.1111/epi.18238

DATA AVAILABILITY STATEMENT

Data from the Norwegian Mother, Father, and Child Cohort Study and the Medical Birth Registry of Norway used in this study are managed by the national health register holders in Norway (Norwegian Institute of Public Health) and can be made available to researchers, provided approval from the Regional Committees for Medical and Health Research Ethics, compliance with the EU General Data Protection Regulation, and approval from the data owners. The consent given by the participants does not apply to storage of data on an individual level in repositories or journals. Researchers who want access to datasets for replication should apply through helsedata.no. Access to datasets requires approval from The Regional Committee for Medical and Health Research Ethics in Norway and an agreement with MoBa.