Long before scientists could read DNA, physicians had noticed that epilepsy sometimes appeared repeatedly within families.
For much of the nineteenth and early twentieth centuries, however, ‘hereditary epilepsy’ was an imprecise idea. Family clustering could be recognised, but there was no way to identify a specific molecular cause, distinguish inherited susceptibility from shared environment, or explain why one relative developed seizures while another did not.
The modern genetics of epilepsy emerged in stages: careful family histories, twin studies, linkage mapping, the discovery of single-gene epilepsies, the realisation that many of the first genes encoded ion channels, and finally large-scale sequencing that revealed a much broader genetic landscape.
This history also contains an important warning. Early ideas about heredity were sometimes absorbed into eugenic policies that treated people with epilepsy as biologically defective. Modern epilepsy genetics grew from scientific efforts to understand disease mechanisms, but it developed against that earlier history of misuse.
Family clustering came before genetics as a laboratory science
Physicians had described familial epilepsy for centuries.
By the nineteenth century, increasingly detailed medical histories made it clear that some forms of epilepsy occurred more often among relatives than chance alone would seem to explain.
But the word hereditary covered many different possibilities.
A family might contain several people with seizures caused by one highly penetrant gene. Another family might carry many genetic susceptibility variants, each contributing only a small amount of risk. Shared infections, injuries, living conditions or diagnostic habits could also make disease appear familial.
Without molecular tools, these possibilities could not be separated.
Mendelian genetics changes the language of inheritance
Gregor Mendel’s work on inheritance was rediscovered around 1900, helping establish the idea that some traits could pass through families according to predictable rules.
Medical researchers began applying Mendelian reasoning to human disorders.
Epilepsy was difficult.
Some rare familial seizure disorders appeared compatible with dominant or recessive inheritance, but most common epilepsies did not follow simple Mendelian patterns.
Variable expression, incomplete penetrance and uncertain diagnosis made pedigree interpretation difficult.
The dangerous detour into eugenics
In the early twentieth century, the observation that epilepsy could run in families was sometimes transformed into sweeping claims that epilepsy represented hereditary ‘degeneration’.
Those claims were scientifically crude.
They helped support marriage restrictions, institutional segregation and forced sterilisation laws in several countries.
The later development of molecular epilepsy genetics demonstrated just how misleading the old category of ‘hereditary epilepsy’ had been. Epilepsy is not one inherited disorder. It is a large group of conditions with many different genetic, structural, metabolic, infectious, immune and unknown causes.
The site’s separate history pages examine the eugenics and discriminatory-law era in detail.
William Lennox and the ‘testimony of twins’
One of the major steps toward scientific epilepsy genetics came from twin studies.
William Gordon Lennox recognised that identical and non-identical twins offered a natural way to estimate genetic influence.
Monozygotic twins share essentially the same inherited DNA sequence, while dizygotic twins share, on average, about half of their segregating genetic variation.
If a particular epilepsy occurred more often in both members of identical twin pairs than in fraternal pairs, that would support a genetic contribution.
Lennox accumulated twin data from the 1930s onward and later described the importance of the ‘testimony of twins’ in his 1960 work with Margaret Lennox.
Twin studies did not prove one ‘epilepsy gene’
The value of the twin work was subtler.
It showed that heredity mattered much more for some epilepsy types than others.
Later reanalysis of Lennox’s historical records using modern syndrome classifications found particularly strong genetic influence in idiopathic generalised epilepsies.
The same work also showed why epilepsy genetics could not be reduced to one mode of inheritance.
Different epilepsy syndromes showed different patterns of concordance.
From ‘hereditary’ to ‘idiopathic’ epilepsy
Twentieth-century epilepsy classification often used the term idiopathic for epilepsies that appeared to arise without a visible structural lesion and were presumed to have an inherited tendency.
The word did not mean that a specific gene had been identified.
Instead, it often implied that epilepsy itself was the primary disorder rather than a consequence of an acquired brain lesion.
As molecular discoveries accumulated, this terminology would eventually be revised because ‘idiopathic’ could conceal very different genetic mechanisms.
Families became maps
Before whole-genome sequencing was possible, geneticists used large pedigrees to locate disease genes indirectly.
The key technique was linkage analysis.
Researchers followed inherited DNA markers through large families and asked which chromosomal regions consistently travelled with the epilepsy phenotype.
If affected relatives repeatedly inherited the same marker region, the disease-causing gene was likely to lie nearby.
Linkage could identify a chromosomal neighbourhood long before the gene itself was known.
Why rare large families mattered so much
Common epilepsies were difficult to map because their inheritance was often complex.
Rare families with many affected relatives were more informative.
A single large pedigree might contain enough meioses — enough transmissions from parent to child — to reveal a clear linkage signal.
This created an important historical pattern: the first epilepsy genes were discovered not because the disorders were common, but because their inheritance was genetically tractable.
Benign familial neonatal seizures point toward chromosome 20
Benign familial neonatal convulsions — now generally referred to using more modern terminology such as self-limited familial neonatal epilepsy — provided one of the early successful linkage stories.
Families showed autosomal dominant inheritance of seizures beginning in the first days of life.
Linkage studies mapped one form to chromosome 20 and another to chromosome 8.
These findings demonstrated that a clearly defined epilepsy syndrome could be tied to a specific chromosomal region even before its molecular mechanism was known.
Autosomal dominant nocturnal frontal lobe epilepsy
Another decisive family-based discovery involved autosomal dominant nocturnal frontal lobe epilepsy, historically abbreviated ADNFLE.
The syndrome was recognised in large families whose members experienced brief, often dramatic nocturnal seizures.
In 1995, linkage analysis placed a disease locus in a large Australian family on chromosome 20q13.2.
That region contained a biologically plausible candidate: CHRNA4, which encodes the alpha-4 subunit of a neuronal nicotinic acetylcholine receptor.
1995: CHRNA4 and the first widely recognised epilepsy gene
In 1995, Ortrud Steinlein and colleagues identified a missense mutation in CHRNA4 that segregated with autosomal dominant nocturnal frontal lobe epilepsy in the large pedigree.
This is commonly described as the first identification of a gene whose mutation directly causes an idiopathic epilepsy syndrome.
The discovery was historically transformative.
For the first time, a human epilepsy could be traced from clinical syndrome to pedigree, chromosome, gene and specific molecular variant.
The first discovery was a receptor-channel gene
CHRNA4 encodes part of a ligand-gated ion channel.
That immediately suggested a mechanistic link between inherited epilepsy and neuronal excitability.
If a mutation altered the behaviour of an ion channel expressed in the brain, it could change the probability that neurons would fire synchronously and generate seizures.
This helped launch the idea that inherited epilepsies might often be channelopathies.
1998: KCNQ2 and KCNQ3
The channelopathy idea strengthened dramatically in 1998.
Researchers identified mutations in KCNQ2 and KCNQ3 in families with familial neonatal seizures.
These genes encode subunits of neuronal potassium channels that contribute to the M-current, an important brake on repetitive neuronal firing.
Reducing that stabilising current made a direct biological connection between altered membrane excitability and inherited seizure susceptibility.
The discovery provided one of the clearest early examples of molecular genetics explaining why an apparently normal newborn brain could nevertheless enter recurrent seizures.
The ‘channelopathy era’
During the late 1990s and early 2000s, several major epilepsy discoveries involved genes encoding ion channels or neurotransmitter receptors.
This produced an understandable enthusiasm for the idea that genetic epilepsy was principally a disorder of channels.
Voltage-gated sodium and potassium channels, nicotinic acetylcholine receptors and GABA-A receptor subunits all became implicated.
For a time, the field seemed to be converging on a simple model: epilepsy genes would be genes controlling membrane excitability.
That model was important — but incomplete.
GEFS+ shows that one family can contain several seizure phenotypes
Generalised epilepsy with febrile seizures plus — GEFS+ — complicated the idea that one mutation should produce one uniform syndrome.
Within the same extended family, different people could have febrile seizures, febrile seizures persisting beyond the usual age range, generalised tonic-clonic seizures and other seizure types.
Mutations in genes including SCN1B, SCN1A and GABRG2 were identified in some GEFS+ families.
The lesson was crucial: a shared molecular cause could produce a spectrum of epilepsy phenotypes.
SCN1A becomes one of the central epilepsy genes
SCN1A encodes the alpha subunit of the Nav1.1 voltage-gated sodium channel.
Mutations were first linked to GEFS+ families around 2000.
Then, in 2001, de novo SCN1A mutations were identified in children with severe myoclonic epilepsy of infancy, the syndrome now widely known as Dravet syndrome.
This was a major conceptual advance.
The child did not need a multigenerational family history for epilepsy to be genetic.
A disease-causing variant could arise newly in the affected child.
‘Genetic’ no longer meant ‘inherited from a parent’
The discovery of de novo mutations changed clinical thinking.
For families, the absence of epilepsy in parents or siblings had often been taken as evidence against a genetic cause.
Molecular diagnosis showed that this was wrong.
A child’s epilepsy could be genetic because of a newly arisen pathogenic variant that was not present constitutionally in either parent.
This distinction remains fundamental to modern genetic counselling.
Dravet syndrome changed more than diagnosis
The link between SCN1A and Dravet syndrome helped transform a clinical syndrome into a molecularly defined disease spectrum.
It also showed that severe developmental epilepsy could arise from dysfunction of one ion-channel gene even when early brain imaging appeared structurally normal.
Later research would reveal a wider range of SCN1A-related phenotypes and more complex genotype–phenotype relationships.
GABRG2 broadens the mechanism
In 2001, mutations in GABRG2, encoding a gamma subunit of the GABA-A receptor, were linked with epilepsy phenotypes including childhood absence epilepsy and febrile seizures.
This strengthened the idea that genetic epilepsy could arise not only from voltage-gated channels but also from altered inhibitory neurotransmission.
Excitation and inhibition were becoming molecular genetic concepts rather than only physiological ones.
2002: LGI1 breaks the channel-only model
A major challenge to the channelopathy model came with the discovery that mutations in LGI1 cause autosomal dominant lateral temporal epilepsy.
LGI1 does not encode an ion channel.
It encodes a secreted protein involved in synaptic organisation and signalling.
This discovery showed that genetic epilepsy could arise from defects in synaptic biology that were not themselves channel mutations.
The field began widening beyond membrane excitability toward neuronal development, synaptic organisation and network function.
One gene, many phenotypes; one phenotype, many genes
As more families were studied, a simple one-gene/one-syndrome model became increasingly difficult to maintain.
The same gene could cause markedly different epilepsy phenotypes.
Conversely, patients with very similar electroclinical syndromes could carry pathogenic variants in completely different genes.
Genetic epilepsy was beginning to look less like a catalogue of isolated disorders and more like a web of overlapping biological pathways.
Association studies promised more than they initially delivered
Rare Mendelian families were relatively tractable.
Common epilepsies were much harder.
During the 1990s and 2000s, researchers performed many candidate-gene association studies, often testing variants in ion-channel genes because those genes seemed biologically plausible.
Many reported associations failed to replicate reliably in independent populations.
This was an important corrective period.
Genetic plausibility was not enough. Large samples, rigorous statistics and replication were essential.
Copy-number variation adds another genetic mechanism
Not all disease-causing genetic changes alter a single DNA letter.
Some delete or duplicate larger stretches of chromosome.
In 2009, recurrent microdeletions at chromosome 15q13.3 were shown to increase risk of idiopathic generalised epilepsy.
The finding was historically important because it demonstrated that structural genomic variation could contribute to common epilepsy susceptibility.
It also highlighted incomplete penetrance: some carriers could remain unaffected.
The Human Genome Project changes the scale
The completion of the Human Genome Project and rapid advances in sequencing technology changed epilepsy genetics from a family-by-family enterprise into a genomic field.
Researchers no longer had to identify a chromosomal region first and then painstakingly sequence candidate genes one at a time.
Increasingly, many genes — eventually all protein-coding genes or the entire genome — could be examined simultaneously.
Next-generation sequencing changes the question
Gene panels and exome sequencing became especially powerful for severe childhood epilepsies.
Instead of asking, ‘Which known syndrome does this child resemble closely enough to justify sequencing one gene?’, clinicians could ask, ‘Which of hundreds of possible epilepsy genes contains the relevant variant?’
This reversed part of the diagnostic process.
Phenotype still mattered, but sequencing could identify causes that clinicians had not predicted from appearance alone.
2013: Epi4K and the era of large collaborative sequencing
In 2013, the Epi4K Consortium and Epilepsy Phenome/Genome Project published a major exome-sequencing study of children with infantile spasms and Lennox–Gastaut syndrome.
By sequencing affected children together with their parents, researchers could identify de novo variants systematically.
The study demonstrated an excess of de novo mutations in genes highly intolerant of functional variation and implicated genes including GABRB3 and ALG13.
The scale of collaboration marked a new era.
Gene discovery increasingly depended on international cohorts rather than one exceptional pedigree.
From hundreds of families to thousands of genomes
This shift was partly statistical.
Many severe epilepsies are individually rare and genetically heterogeneous.
A pathogenic variant may be seen only once or twice in one clinic.
Pooling thousands of cases allows researchers to detect recurrence, establish statistical enrichment and distinguish true disease genes from background human variation.
Modern epilepsy genetics therefore became inseparable from large collaborative databases and shared phenotyping.
The idea of an ‘epilepsy gene’ becomes less simple
The phrase epilepsy gene remains useful shorthand, but it can mislead.
Some genes cause epilepsy as their main clinical effect.
Others cause broader neurodevelopmental disorders in which seizures are only one feature.
Some variants strongly determine disease, while others merely alter susceptibility.
Some pathogenic variants are inherited, others de novo, and still others occur as mosaic mutations in only a proportion of cells.
The field that began by searching for single inherited defects now encompasses an enormous range of genetic architectures.
Genetic and structural epilepsy are not always opposites
Modern discoveries also weakened the historical assumption that an epilepsy was either genetic or structural.
A genetic variant can cause a structural malformation such as focal cortical dysplasia.
Pathogenic variants in pathways including mTOR signalling can produce developmental brain lesions that are both genetic in cause and structural in manifestation.
The old diagnostic boxes were therefore not biologically exclusive.
Mosaicism changes the family-history question again
Some pathogenic variants arise after fertilisation and are present in only a subset of the body’s cells.
This is called mosaicism.
A variant confined largely to brain tissue may be difficult or impossible to detect in a routine blood test.
Parental mosaicism can also affect recurrence risk even when a parent is clinically unaffected.
This added another layer of complexity to the early assumption that inherited disease should be visible in a pedigree.
Common epilepsies remain genetically difficult
Despite spectacular progress in rare monogenic epilepsies, many common epilepsies do not have one identifiable causal gene.
Twin and family studies provide strong evidence of genetic influence, particularly in genetic generalised epilepsies, but much of that risk is distributed across many variants.
This polygenic architecture helps explain why common epilepsy genetics proved so much harder than the first Mendelian successes suggested.
The terminology changes from ‘idiopathic’ to ‘genetic’
As molecular evidence accumulated, epilepsy classification also changed.
The ILAE increasingly moved away from using idiopathic as a broad causal label and adopted genetic as an aetiological category.
Importantly, ‘genetic’ does not require that a pathogenic variant has already been identified.
It can also describe an epilepsy for which family, twin and epidemiological evidence strongly supports a genetic contribution.
This modern usage is much more precise than the older idea of ‘hereditary epilepsy’.
Genetic testing moves from research laboratory to clinic
By the 2010s, genetic testing was increasingly part of clinical epilepsy practice rather than an exceptional research investigation.
Panels, exome sequencing and later genome sequencing allowed molecular diagnoses to be made in children and adults whose epilepsy had remained unexplained for years.
A molecular diagnosis could clarify syndrome, recurrence risk and prognosis and, in some disorders, influence treatment.
The site’s dedicated modern genetic-testing page covers those present-day uses in detail.
The historical reversal
For centuries, physicians started with a seizure phenotype and tried to guess its cause.
Molecular genetics sometimes reversed that direction.
A pathogenic variant could identify a disease mechanism first, after which clinicians recognised that patients carrying variants in the same gene could present with a broader range of seizure types than expected.
Genotype began reshaping phenotype.
What the early pioneers got right — and wrong
Early physicians were right that heredity mattered in epilepsy.
They were wrong when they treated epilepsy as one unitary hereditary disorder.
Twin studies were right that genetic influence differed between epilepsy types.
The first molecular discoveries were right that ion channels were important.
They were wrong only if interpreted to mean that all genetic epilepsies were channelopathies.
Each stage added a real insight while also exposing the limits of the previous model.
The central historical lesson
The modern genetics of epilepsy did not begin with DNA sequencing.
It began with families.
Pedigrees suggested inheritance. Twin studies quantified it. Linkage analysis mapped disease to chromosomes. Molecular genetics identified the first causal genes. Ion-channel discoveries provided mechanisms. De novo mutations showed that genetic did not necessarily mean inherited. Copy-number variants and non-channel genes widened the field. Next-generation sequencing then transformed gene discovery from one pedigree at a time into an international genomic enterprise.
The most important change was conceptual.
The old question was: Is epilepsy hereditary?
The modern question is much more precise: Which genetic mechanism contributes to this particular epilepsy, how strongly does it act, and how does it interact with development, brain structure and the rest of the genome?
Sources and further reading
Poduri A, Lowenstein D. Epilepsy genetics — past, present, and future. Current Opinion in Genetics & Development. 2011.
Vadlamudi L et al. Epilepsy in twins: insights from unique historical data of William Lennox. Neurology. 2004.
Vadlamudi L et al. Genetics of epilepsy: the testimony of twins in the molecular era. Neurology. 2014.
Steinlein OK et al. A missense mutation in the neuronal nicotinic acetylcholine receptor alpha 4 subunit is associated with autosomal dominant nocturnal frontal lobe epilepsy. Nature Genetics. 1995.
Singh NA et al. A novel potassium channel gene, KCNQ2, is mutated in an inherited epilepsy of newborns. Nature Genetics. 1998.
Claes L et al. De novo mutations in the sodium-channel gene SCN1A cause severe myoclonic epilepsy of infancy. American Journal of Human Genetics. 2001.
Wallace RH et al. Mutant GABA-A receptor gamma2-subunit in childhood absence epilepsy and febrile seizures. Nature Genetics. 2001.
Morante-Redolat JM et al. Mutations in the LGI1/Epitempin gene cause autosomal dominant lateral temporal epilepsy. Human Molecular Genetics. 2002.
Helbig I et al. 15q13.3 microdeletions increase risk of idiopathic generalized epilepsy. Nature Genetics. 2009.
Epi4K Consortium and Epilepsy Phenome/Genome Project. De novo mutations in epileptic encephalopathies. Nature. 2013.
Berkovic SF. The Genetics of Epilepsy. Annual Review of Genomics and Human Genetics. 2020.
Historical evidence checked: September 2026.