Breastfeeding is usually possible for people with epilepsy who take antiseizure medicines. The decision is not simply “medicine versus no medicine”. It involves balancing the well-established benefits of breastfeeding, the importance of maintaining seizure control, the amount of medicine transferred into milk, the age and health of the baby, whether one or several medicines are being used, and how well that particular medicine has been studied during lactation.
Current guidance does not support routinely advising people with epilepsy to avoid breastfeeding simply because they take an antiseizure medicine. NICE states that breastfeeding is generally safe for most women and girls taking antiseizure medications and should be encouraged, while decisions about individual medicines should be made with the prescriber using drug-specific information.
Breastfeeding exposure is different from pregnancy exposure
During pregnancy, a fetus may be exposed to a medicine continuously through the placenta. After birth, exposure through breast milk depends on how much of the medicine enters milk, how much milk the baby drinks, how well the medicine is absorbed from the infant gut, and how efficiently the infant can metabolise and eliminate it.
A medicine that carries an important pregnancy risk does not automatically create the same level of risk through breast milk. Conversely, a medicine considered acceptable in pregnancy can still require infant monitoring during breastfeeding.
Most antiseizure medicines do enter breast milk
The important question is not simply whether a medicine can be detected in milk. Many medicines can. The more useful questions are how much reaches the infant, whether infant blood levels are clinically important, and whether adverse effects have actually been observed.
The International League Against Epilepsy Women Task Force reviewed 85 publications on antiseizure medicines during breastfeeding. For medicines with available infant blood-level data, very low infant concentrations — around 10% or less of maternal serum concentrations — were reported for carbamazepine, gabapentin, levetiracetam, oxcarbazepine, phenytoin, valproate and clonazepam. Somewhat higher concentrations have been reported with lamotrigine and topiramate. Higher infant levels have been reported with ethosuximide, phenobarbital and zonisamide. Evidence for many newer medicines remains limited.
Breastfeeding is generally encouraged
The ILAE review concluded that the available evidence does not show an increase in adverse effects large enough to outweigh the recognised benefits of breastfeeding overall. Long-term data are still incomplete, especially for newer antiseizure medicines, but available developmental studies have not shown poorer outcomes in breastfed children exposed to the better-studied medicines.
Large prospective studies measuring paired maternal and infant blood samples have also shown that infant antiseizure medicine concentrations are generally substantially lower than maternal concentrations.
The baby’s age matters
A healthy full-term infant several weeks old usually handles medicines more efficiently than a premature infant or a newborn in the first days of life. Kidney and liver function are still developing after birth, so the same milk concentration can have a different effect depending on gestational age, postnatal age and medical condition.
Extra caution may therefore be needed with premature infants, medically fragile newborns, babies with liver or kidney disease, and babies who are exclusively breastfed while the mother is taking a medicine that has relatively high transfer into milk.
Combination treatment can change the picture
Most breastfeeding safety information comes from relatively small observational studies, and many participants were taking more than one medicine. Polytherapy can complicate interpretation because sedating effects may add together and because some antiseizure medicines alter the metabolism of others.
An infant exposed through breast milk to two or three medicines may therefore need closer observation than an infant exposed to a single well-studied medicine at a stable dose.
Lamotrigine
Lamotrigine is one of the most commonly discussed antiseizure medicines during breastfeeding because it can pass into breast milk in clinically measurable amounts. Infant concentrations are usually lower than maternal concentrations, but higher infant exposure than with some older medicines has been reported.
Most breastfed infants exposed to lamotrigine do well. However, occasional reports have described drowsiness, poor sucking, rash, breathing problems and other adverse effects. LactMed recommends watching for symptoms such as rash, apnoea, excessive sleepiness or poor feeding, particularly in younger infants or when maternal doses are high.
A major postpartum issue is that lamotrigine clearance can fall rapidly after delivery. If the dose was increased during pregnancy, maternal blood concentrations can rise after birth unless the dose is brought back toward the pre-pregnancy plan. That can increase maternal side effects and potentially increase transfer into milk. This is why antenatal planning for postpartum dose reduction is important.
Levetiracetam
Levetiracetam can appear in breast milk in appreciable concentrations, but infant blood levels are usually much lower than maternal levels. Most reports have not found serious problems.
Occasional infant sedation has been reported. Monitoring for unusual sleepiness, feeding difficulty, adequate weight gain and developmental progress is sensible, particularly in young or exclusively breastfed infants and when levetiracetam is combined with other medicines.
Like several other antiseizure medicines, levetiracetam levels can change substantially during pregnancy and again after delivery, so postpartum maternal dose review may be important.
Carbamazepine
Carbamazepine has one of the longer breastfeeding safety records. Breastfed infants usually have blood concentrations below the therapeutic range used to treat epilepsy, and most infants have no adverse effects.
Rare reports have described drowsiness, poor sucking, poor weight gain and liver abnormalities. Infants should be observed for jaundice, feeding problems and excessive sleepiness, particularly during the early neonatal period and where other medicines are also being used.
Valproate
Valproate concentrations in breast milk are generally low and infant blood concentrations are usually undetectable or low. Breastfeeding itself is not normally a reason to stop valproate if the medicine is clinically required.
Rare or theoretical concerns include liver toxicity and abnormal bruising or bleeding. Monitoring for jaundice, unusual bruising, bleeding, feeding difficulty or marked drowsiness is appropriate.
This is separate from the much more serious reproductive safety issues associated with valproate exposure during pregnancy. UK prescribing restrictions and Pregnancy Prevention Programme requirements must still be followed for people who could become pregnant.
Phenytoin
Phenytoin transfer into breast milk is generally low because of its protein binding. It has a long history of use during breastfeeding, and clinically important infant exposure is uncommon. Monitoring remains appropriate if other sedating medicines are also being taken or if the infant is premature or medically vulnerable.
Oxcarbazepine
Available evidence suggests relatively low infant exposure to oxcarbazepine and its active metabolite, but the evidence base is smaller than for carbamazepine. No major pattern of adverse developmental outcome has been established from the limited available data.
Gabapentin
Gabapentin is transferred into milk, but infant serum concentrations reported in the ILAE review were generally low. Published experience is still much smaller than for older antiseizure medicines. Observation for excessive sleepiness, poor feeding and poor weight gain is reasonable.
Topiramate
Topiramate can reach measurable infant concentrations and transfer through breast milk may be greater than with several older medicines. Most reported infants have remained well, but diarrhoea and sedation have been described in individual reports.
Topiramate also has important pregnancy-related restrictions in the UK because of fetal and neurodevelopmental risk. Those pregnancy restrictions should not be confused with breastfeeding exposure, but they remain highly relevant to contraception and future pregnancy planning.
Phenobarbital and primidone
Phenobarbital can accumulate in young infants because newborn elimination is slow. Sedation and poor feeding are the main concerns. Primidone is partly converted to phenobarbital and can create similar issues.
If an infant becomes unusually sleepy, feeds poorly or fails to gain weight, urgent clinical review is appropriate. Abrupt cessation of breastfeeding after prolonged exposure to sedating antiseizure medicines can also theoretically produce withdrawal symptoms in the infant, so changes should be discussed with the clinical team.
Ethosuximide
Ethosuximide can produce relatively high infant exposure compared with many other antiseizure medicines. The evidence base is limited, and closer infant monitoring may be appropriate where it is used during breastfeeding.
Zonisamide
Zonisamide can produce comparatively high infant serum concentrations and has a smaller breastfeeding evidence base. This does not mean breastfeeding is automatically impossible, but specialist individual assessment is particularly important.
Newer antiseizure medicines: evidence can be sparse
For several newer or less commonly used medicines, direct evidence on breast-milk transfer, infant blood concentrations and long-term developmental outcomes remains limited or absent. The ILAE review found no milk or infant-level data for several medicines, including cenobamate, clobazam, eslicarbazepine acetate, cannabidiol, rufinamide, stiripentol, tiagabine and vigabatrin at the time of review.
Limited evidence does not automatically mean a medicine is unsafe. It means uncertainty is greater and decisions need to be more individualised.
What symptoms should be watched for?
Most breastfed infants exposed to antiseizure medicines do not develop serious adverse effects. Nevertheless, parents and clinicians should know what to watch for.
- unusual or increasing drowsiness
- difficulty waking for feeds
- poor sucking or reduced feeding
- poor weight gain
- repeated vomiting or diarrhoea
- rash
- breathing pauses or unusual breathing
- jaundice
- unusual bruising or bleeding
- reduced alertness or unusual behaviour
These symptoms have many possible causes and do not automatically mean antiseizure medicine toxicity, but they justify clinical assessment.
Routine infant blood tests are not always necessary
Most healthy infants do not need routine antiseizure medicine blood levels simply because they are breastfed. Testing may be useful when the infant develops concerning symptoms, when exposure is expected to be relatively high, when the maternal dose is high or changes substantially, or when the infant is premature or medically vulnerable.
The specific medicine matters. For example, serum levels may be useful if lamotrigine toxicity is suspected.
Maternal dose changes after birth can matter as much as milk transfer
Pregnancy can increase the clearance of medicines such as lamotrigine and levetiracetam, leading clinicians to increase doses to maintain seizure control. After delivery, clearance may return toward the pre-pregnancy state quickly.
NICE advises that when antiseizure medicine doses have been changed during pregnancy, an antenatal plan should be made to return toward pre-conception doses, beginning in the first few days after birth. This protects the mother from excessive drug exposure and can also reduce unnecessary infant exposure through breast milk.
Do not reduce or stop epilepsy medicine just to breastfeed
Stopping or reducing antiseizure medicine without specialist advice can allow seizures to recur. The postnatal period can already contain important seizure triggers, particularly severe sleep deprivation and missed doses.
A medication change intended to reduce infant exposure can therefore create a greater overall risk if it destabilises seizure control.
Sleep deprivation is a separate postnatal safety issue
Feeding a newborn can cause major sleep disruption. For people whose seizures are triggered by sleep deprivation, the feeding plan may need to include protected sleep periods, expressed milk, shared night-time care or combination feeding where appropriate.
The safest feeding arrangement is not necessarily the one that maximises exclusive breastfeeding at all costs. Maternal seizure control, infant safety and sustainable sleep are part of the same clinical picture.
Breastfeeding and rescue medication
Occasional emergency treatment for seizures does not automatically mean breastfeeding must stop permanently. However, sedating rescue medicines can affect maternal alertness and may enter milk to different degrees. Advice depends on the drug, dose, frequency and infant age.
After a seizure requiring rescue medication, another adult may need to supervise infant care until the mother is fully alert and physically recovered.
Breastfeeding after a seizure
A seizure itself does not contaminate breast milk. The main concerns after a seizure are the parent’s recovery, injuries, awareness, medication given during the event and whether it is safe to hold or feed the baby without assistance.
If awareness is impaired, feeding should wait until the person has recovered enough to hold and respond to the infant safely.
Safety during feeds matters
For someone at risk of sudden loss of awareness, falls or bilateral tonic-clonic seizures, practical adaptations can reduce injury risk. Feeding while seated low down, avoiding feeding on a high chair or sofa edge, and having another adult nearby during higher-risk periods may be useful.
These are practical safety adaptations, not reasons to discourage breastfeeding.
Long-term child development
Long-term evidence is much smaller than short-term exposure data. Available studies of children whose mothers used medicines including carbamazepine, lamotrigine, levetiracetam, phenytoin and valproate have not shown worse developmental outcomes attributable to breastfeeding.
Some cohorts have actually reported equal or better developmental or cognitive outcomes among breastfed children, but these observational findings do not prove that antiseizure medicine exposure through breast milk improves development. They mainly provide reassurance that breastfeeding exposure has not shown a clear developmental harm signal in the better-studied medicines.
Why individual counselling matters
There is no single breastfeeding rule that applies equally to every antiseizure medicine, every dose and every baby.
A useful discussion includes the exact medicine and dose, whether treatment is monotherapy or polytherapy, pregnancy-related dose changes, infant gestational age, infant health, whether feeding is exclusive or partial, maternal seizure control, sleep deprivation risk and what signs should prompt review.
The central message
For most people with epilepsy, antiseizure medicine treatment is not a reason to avoid breastfeeding. The evidence supports breastfeeding with most commonly used antiseizure medicines, provided decisions are individualised and infant monitoring is matched to the medicine and clinical situation.
The most important mistake is treating all antiseizure medicines as though they behave the same way. They do not. Transfer into breast milk, infant blood levels, sedation risk and the strength of available evidence vary considerably between drugs.
Equally, breastfeeding advice should never ignore the parent’s epilepsy. Maintaining seizure control, avoiding abrupt medication changes and protecting sleep are essential parts of postnatal care.
Related pages: Hormones, Contraception & Pregnancy · Treatment Options · Medication Side Effects and Interactions · Sleep & Epilepsy · Epilepsy Care Plans
Sources and further reading
NICE NG217 — Epilepsies in children, young people and adults: treatment, safety, monitoring and withdrawal
Tomson T et al. Breastfeeding while on treatment with antiseizure medications: a systematic review from the ILAE Women Task Force. Epileptic Disorders. 2022.
ILAE Women Task Force systematic review — full text
Shawahna R, Zaid L. Concentrations of antiseizure medications in breast milk of lactating women with epilepsy: systematic review. Seizure. 2022.
Birnbaum AK et al. Antiepileptic Drug Exposure in Infants of Breastfeeding Mothers With Epilepsy. JAMA Neurology. 2020.
Breastfeeding recommendations for women taking anti-seizure medications. Epilepsy & Behavior. 2022.
LactMed — Lamotrigine
LactMed — Levetiracetam
LactMed — Carbamazepine
LactMed — Valproic Acid
LactMed — Lacosamide
MHRA — Antiepileptic drugs in pregnancy: updated safety advice
Evidence checked: September 2026.