A sudden jerk of your toddler’s arm during a feverish night can feel terrifying. The same word—“seizure”—gets whispered in the ER, yet the chart may later read “febrile spasm,” leaving parents unsure what actually happened.
Understanding the real boundary between seizure and spasm changes triage timing, medication choices, and long-term expectations. Below, every distinction is unpacked with the clinical nuance families and frontline clinicians need at 2 a.m.
Core Definitions and Neural Origins
A seizure is an uncontrolled surge of electrical activity across cortical neurons that recruits adjacent or distant brain regions. The event can be focal, starting in one lobe, or generalized from onset, and always involves detectable EEG changes.
Spasms are brief, symmetric or asymmetric muscular contractions driven by a subcortical trigger—often brainstem or spinal—without the cascading cortical rhythm that defines epilepsy. They show no ictal EEG correlate unless a concurrent seizure is present.
This neuroanatomic divide explains why anticonvulsants that silence glutamate storms do little for infantile spasms rooted in aberrant circuitry below the cortex.
Semiology in Real Life
Visual Signatures
Seizures present as rhythmic twitching that accelerates then slows, lip-smacking, or a blank stare that resists calling. Spasms look like a startle—head drops, arms fling outward, trunk flexes—then vanish in under two seconds.
Recording the event on a phone in landscape mode at 60 fps captures subtle timing cues clinicians can parse frame-by-frame. Slow-motion review often reveals whether the movement is repeating every 200 ms (seizure) or a single lightning-fast flexion (spasm).
Clustering Patterns
Infantile spasms typically cluster on waking, with 10–100 individual spasms in a five-minute shower. Seizures are more stochastic, appearing during sleep, play, or fever, rarely repeating more than twice within an hour unless status develops.
Parents can log clusters on a printed calendar taped to the fridge; the emerging histogram often decides whether urgent neuroimaging or empirical steroids are warranted.
EEG and the Gold-Standard Divide
Hypsarrhythmia—chaotic, high-voltage slow waves with multifocal spikes—flags infantile spasms even when the child looks asymptotic between jerks. Seizures instead show rhythmic spike-and-wave at 3 Hz or faster, evolving in frequency and field.
A 30-minute awake-and-sleep EEG is insufficient; capture the waking-up epoch because hypsarrhythmia may vanish once drowsy. If suspicion stays high, admit for overnight video-EEG rather than repeating short outpatient studies.
Age as a Diagnostic Lens
Spasms peak between 3 and 9 months, seldom appear after 18 months, and almost never start after age three. Seizures can debut from premature neonates to nonagenarians, making age an instant sieve.
A previously normal 14-month-old who suddenly flexes at the waist twice daily is statistically more likely to have seizures mimicking spasms than true infantile spasms, guiding the differential toward genetic epilepsies.
Fever’s Dual Role
Febrile Spasms
Simple febrile seizures last <15 min, occur once in 24 h, and lack focal features; they are seizures, not spasms. Complex febrile events breach one of those limits yet still show rhythmic semiology.
True febrile spasms do not exist in taxonomy—if fever accompanies flexor episodes, suspect meningitis or metabolic crisis rather than reclassifying the movement.
Hot Bath Test
Some infants have heat-sensitive spasms that erupt during warm baths; the same child may stay spasm-free if water is kept at 36 °C. This observation can spare empiric steroids when the EEG is borderline.
Metabolic Masqueraders
Pyridoxine-dependent epilepsy causes seizures that vanish within hours of 100 mg IV pyridoxine, whereas infantile spasms from Menkes disease may transiently worsen after the same dose. Glucose transporter-1 deficiency triggers seizures on fasting, but spasms are not part of its core phenotype.
Order serum lactate, ammonia, and CSF glucose simultaneously; a low CSF glucose (<40 mg/dL) with normal blood sugar points to GLUT-1DS, shifting therapy from steroids to ketogenic diet.
Pharmacologic Response Profiles
Adrenocorticotropic hormone (ACTH) stops infantile spasms in 70 % within two weeks, yet fails in most focal-onset seizures. Vigabatrin works best for tuberous-sclerosis-related spasms but can exacerbate absence seizures.
Conversely, levetiracetam quiets many generalized seizures yet can provoke increased spasm frequency in West syndrome. Document baseline counts before any new drug so directionality of change is clear.
Imaging Clues
Malformations of Cortical Development
Tuberous-sclerosis tubers are hypointense on T1, hyperintense on FLAIR, and often sit next to ventricles; when spasms emerge, look for subependymal giant-cell astrocytoma that can obstruct foramina. Focal cortical dysplasia type IIa shows transmantle sign—linear tapering from cortex to ventricle—best seen on 3-T MRI with 1 mm slices.
Diffusion in Acute Setting
Status-epilepticus can create transient cortical diffusion restriction that mimics stroke; spasms never do. If DWI brightens after prolonged seizure, repeat MRI in four weeks to confirm resolution and avoid mislabeling as ischemia.
Genetic Landscape
ARX mutations yield X-linked infantile spasms with dystonia, absent corpus callosum, and microcephaly. CDKL5 disorder starts with seizures at 6 weeks, evolving into spasms by 3 months, then Lennox-Gastaut features by age two.
Targeted epilepsy gene panels now include 500+ genes; obtain trio exome if no structural lesion is found because precision therapy—such as mTOR inhibitors for TSC2—depends on exact variant.
Parental Action Plan
First-Aid Kit
Keep a timer, a printed seizure action plan, and 5 mg midazolam buccal syringes in the diaper bag. Note exact clock time when the episode starts; ER staff convert to video timestamp if footage is provided.
Red-Flag Calendar
Mark any day with >20 spasms, focal features, or post-ictal sleep >30 min. Three red days in a week justifies urgent re-evaluation regardless of scheduled follow-up.
Long-Term Outlook
Spasms controlled before 4 months of age yield normal cognition in 40 % of cryptogenic cases; delay beyond 12 months drops that figure below 10 %. Early seizure freedom in Dravet syndrome does not prevent later developmental plateau, so expectations must be syndrome-specific.
Schedule neuropsychology at 12 months even if spasms vanish; subtle visuospatial deficits emerge early and respond to targeted therapy if caught before preschool.
Quality-of-Life Tactics
Shift bedtime 30 min earlier during ACTH course; steroid insomnia peaks at week two and can be halved with melatonin 1 mg given at 6 p.m. Use a weighted 2 kg blanket for sensory input if spasms leave the child irritable, but discontinue at once if any apnea is noted.
Join virtual support groups stratified by etiology—tuberous-sclerosis parents trade MRI timing tips, while GLUT-1 families swap ketogenic recipes. This specificity prevents the anxiety spillover that comes from mixing prognosis categories.
Clinical Pearls for Professionals
Always re-examine the EEG after ACTH taper; hypsarrhythmia can re-emerge silently, and second remission is easier to achieve than first. Document baseline visual fields before vigabatrin using Goldman perimetry; children under 3 can cooperate if the test is gamified with sticker rewards every 5 min.
When parental video conflicts with your bedside exam, trust the timestamped footage; observers misclassify spasms as hiccups 15 % of the time. Finally, never label a movement “benign” until both ictal EEG and MRI are reviewed—retraction of a casual phrase later erodes caregiver trust more than an initial honest uncertainty.