Two people can experience apparently similar brain injuries and have very different long-term outcomes. One may never have another neurological event. Another may develop recurrent unprovoked seizures months or years later. The difference is not explained by one single factor.
Post-injury epilepsy risk depends on the type and severity of injury, which brain regions are affected, whether there is bleeding or cortical damage, age, genetics, inflammation, early seizures, network reorganisation and many other biological variables. This page focuses on why risk differs between people and injuries. The broader cellular process by which a brain becomes epileptic is covered separately under Epileptogenesis.
Brain injury is a risk factor, not a guarantee
Traumatic brain injury, stroke, infection, hypoxic–ischaemic injury and prolonged seizures can all increase the risk of later epilepsy. But most people with a brain insult do not inevitably develop recurrent unprovoked seizures.
This is one of the central problems in epilepsy research: clinicians can identify groups at higher risk, but predicting the outcome for one individual remains difficult.
Severity matters
More severe injuries are generally associated with higher later epilepsy risk than milder injuries. In traumatic brain injury, penetrating injuries, major cortical contusions, intracranial haemorrhage and injuries requiring neurosurgery tend to carry more risk than uncomplicated mild concussion.
Severity is not a single measurement, however. Glasgow Coma Scale scores, duration of loss of consciousness, post-traumatic amnesia, imaging findings and the presence of bleeding each describe different parts of the injury.
Where the injury occurs matters
Injury involving cerebral cortex is especially relevant because cortical networks generate many focal seizures. Damage affecting temporal, frontal or other cortical regions can alter local circuitry and long-range connectivity in ways that may support epileptogenesis.
Deep subcortical lesions can also be associated with epilepsy, particularly when they disrupt connected cortical networks. Location is therefore about network anatomy, not simply which lobe contains the visible lesion.
Bleeding can increase risk
Intracranial haemorrhage exposes brain tissue to blood products that can trigger inflammation, iron deposition, oxidative stress and tissue injury. Subdural, subarachnoid, intracerebral and haemorrhagic contusion patterns do not carry identical risk.
Blood–brain barrier disruption may also contribute to later hyperexcitability. Experimental work suggests that proteins entering brain tissue after barrier breakdown can alter astrocyte function, potassium regulation and inflammatory signalling.
Cortical scarring and tissue loss
Areas of tissue loss, gliosis and cortical scarring can become part of an epileptogenic network. But a scar visible on MRI is not automatically the seizure focus, and some people with substantial structural damage never develop epilepsy.
What matters is how surviving networks reorganise around the injury. Abnormal connectivity, altered inhibition and changes in intrinsic excitability can be more important than lesion size alone.
Early seizures after injury
Seizures occurring soon after an acute brain injury are often classified as acute symptomatic seizures. They can result directly from swelling, bleeding, metabolic disturbance or acute tissue damage.
Early seizures can be associated with a higher risk of later epilepsy in some injury groups, but they are not identical to post-traumatic epilepsy. A person can have an acute symptomatic seizure and never develop recurrent unprovoked seizures.
Late seizures carry different meaning
A seizure occurring after the acute injury period is more concerning for an enduring epileptogenic change. In traumatic brain injury research, seizures occurring more than about a week after injury are generally considered late post-traumatic seizures rather than acute symptomatic events.
Exact definitions vary by context and guideline, so the timing should always be interpreted alongside the nature of the injury and clinical evidence.
Age changes vulnerability
The developing and ageing brains respond differently to injury. Children may show substantial plasticity, but developmental stage can also alter how damaged circuits reorganise. Older adults may have additional vascular disease, neurodegeneration or reduced network reserve that changes risk.
Age therefore modifies both the biology of recovery and the background probability of epilepsy.
Genetics may influence who develops epilepsy
Genetic background affects ion channels, inflammatory responses, neuronal survival, synaptic plasticity and repair. It is therefore biologically plausible that two people with similar injuries can have different epileptogenic responses because their underlying molecular systems differ.
Research has identified candidate genetic associations with post-traumatic epilepsy, but there is no routine genetic test that can currently predict with certainty who will develop epilepsy after injury.
Inflammation is variable between people
Brain injury triggers microglial activation, cytokine release, astrocyte responses and changes in the blood–brain barrier. The intensity and duration of this inflammatory response can vary substantially.
Some inflammatory signalling supports healing; other patterns may promote persistent hyperexcitability. Individual immune responses may therefore contribute to different long-term seizure risks even after apparently similar injuries.
Network connectivity before the injury
No two brains have identical connectivity. Pre-existing differences in white-matter organisation, cortical architecture and network efficiency may influence how an injury is absorbed and how abnormal activity spreads afterwards.
An injury affecting a highly connected hub may have a very different effect from a similarly sized lesion in a less connected area.
Network repair can be adaptive or maladaptive
The brain attempts to repair and compensate after injury. New connections form, surviving neurons change receptor and ion-channel expression, glial cells remodel the environment and networks reroute activity.
Most of these processes are intended to restore function. But some forms of plasticity can accidentally create recurrent excitatory loops or weaken inhibitory control. Epileptogenesis can therefore emerge from the same plasticity mechanisms that normally support recovery.
Why lesion size alone is not enough
A large lesion does not automatically mean a high seizure burden, and a small lesion can sometimes sit in a highly epileptogenic location. MRI shows structure, but epilepsy depends on function and connectivity as well.
This is why EEG, semiology, PET, SPECT, MEG or intracranial recordings may add important information when structural imaging alone does not explain seizures.
Why some people have a long latent period
Epilepsy can appear months or years after a brain injury. During this interval, inflammatory signalling, gene expression, synaptic reorganisation, gliosis and changes in inhibition may continue evolving.
The first recognised seizure therefore does not necessarily mark the beginning of the biological process. The epileptogenic network may have been developing silently for a long time.
Traumatic brain injury is not one condition
‘TBI’ includes very different injuries: mild concussion, diffuse axonal injury, focal contusion, penetrating trauma, subdural haematoma and major multi-compartment haemorrhage. Combining all of these under one label can hide major differences in epilepsy risk.
Risk estimates should therefore always be interpreted in relation to the population studied and the injury definition used.
Stroke follows a similar principle
After stroke, later epilepsy risk is influenced by cortical involvement, stroke severity, haemorrhage, younger age and early seizures, among other factors. Haemorrhagic and cortical strokes generally carry greater risk than small deep infarcts.
Again, the underlying principle is network disruption: injuries affecting cortex and causing substantial tissue damage are more likely to create enduring epileptogenic conditions.
Infection adds inflammatory and immune mechanisms
Encephalitis, meningitis and some parasitic infections can damage brain tissue directly while also producing strong inflammatory responses. Residual scars, calcifications, hippocampal injury or immune-mediated changes may contribute to later seizures.
The risk depends on the organism or immune process, the regions affected, illness severity and the extent of lasting structural damage.
Why prevention remains difficult
Antiseizure medicines can reduce early seizures after some brain injuries, but preventing early seizures is not the same as preventing later epilepsy. Trials have not yet established a broadly effective treatment that reliably stops post-injury epileptogenesis in humans.
The challenge is partly predictive: an antiepileptogenic treatment would ideally be given to people who are genuinely at high risk, while avoiding unnecessary treatment in those who would never develop epilepsy.
The search for biomarkers
Researchers are studying EEG changes, high-frequency oscillations, MRI and diffusion measures, inflammatory markers, blood biomarkers and network-connectivity patterns as possible predictors of epileptogenesis.
At present, no single marker can reliably tell an individual patient, ‘you will develop epilepsy’ or ‘you will not’ after a brain injury.
Risk is probabilistic, not destiny
Population studies can identify factors associated with higher or lower epilepsy risk, but they cannot determine an individual outcome with certainty. A high-risk injury may never be followed by epilepsy, while epilepsy occasionally develops after an injury that initially appeared relatively modest.
This uncertainty should be stated plainly. Risk factors change probability; they do not function as guarantees.
What this page does — and does not — explain
This page explains why apparently similar brain injuries can lead to different epilepsy outcomes. The molecular transformation itself is covered on Epileptogenesis: How a Brain Can Become Epileptic, while dedicated future pages on post-traumatic, post-stroke and infection-related epilepsy will examine those conditions in greater clinical depth.
Sources
Jasper’s Basic Mechanisms of the Epilepsies, 5th ed. — Epileptogenesis
Jasper’s Basic Mechanisms of the Epilepsies, 5th ed. — EEG Biomarkers of Epileptogenesis
Jasper’s Basic Mechanisms of the Epilepsies, 5th ed. — BDNF/TrkB Signaling and Epileptogenesis Caused by Trauma
Jasper’s Basic Mechanisms of the Epilepsies, 5th ed. — Prophylaxis of Epileptogenesis in Injury and Genetic Epilepsy Models
Evidence checked: September 2026.