Levetiracetam

What it is

Levetiracetam is a synthetic single-enantiomer molecule, a pyrrolidone derivative, supplied as plain and extended-release tablets, an oral solution and an intravenous injection. It is a white to off-white crystalline powder with a faint odour and a bitter taste, and it is very soluble in water - about 104 g per 100 mL - which is unusual among antiseizure drugs and is why an intravenous form exists. It is chemically unrelated to the older antiepileptic drugs. It was first approved in the United States in 1999 and a 2026 review reports it accounts for roughly 56% of US epilepsy prescriptions.

What the research says

It reduces seizures, and the evidence for that in drug-resistant focal epilepsy as an add-on is solid: in the pivotal placebo-controlled trials it cut weekly partial seizure frequency by 26.1% to 30.1% over placebo. Where the evidence turned out worse than assumed is as a first-choice drug. The independent SANAD II trials, funded by the UK NIHR rather than industry, tested it head-to-head over years and it failed non-inferiority against lamotrigine in focal epilepsy (990 patients) and against valproate in generalised epilepsy (520 patients), with more adverse reactions in both. Its distinctive harm is behavioural: non-psychotic behavioural symptoms in 13% of adults and 38% of children against 6% and 19% on placebo.

Evidence grade: Well established.

How it works

Drug class: Antiseizure medication binding synaptic vesicle protein 2A (SV2A); a pyrrolidone derivative chemically unrelated to older antiepileptic drugs

Levetiracetam does not work the way older antiseizure drugs do. It binds to a protein called SV2A that sits on the membrane of the little sacs (synaptic vesicles) nerve cells use to release their chemical messengers, and SV2A is thought to help regulate how those sacs empty. The label is careful: the molecular significance of that binding is not understood, but across levetiracetam and related molecules the strength of SV2A binding tracked how well they stopped seizures in mice, which is why the SV2A interaction is thought to be the mechanism. (Source 1)

What it is used for

• Effectiveness was established in three randomised double-blind placebo-controlled trials in 904 patients with refractory focal seizures. In the first, levetiracetam cut weekly partial seizure frequency by 26.1% (1000 mg/day) and 30.1% (3000 mg/day) over placebo, both statistically significant. Evidence: established. (Source 2)
• SANAD II randomised 990 people and followed them two years further. Levetiracetam failed non-inferiority against lamotrigine for time to 12-month remission (HR 1.18, 97.5% CI 0.95-1.47), was inferior per-protocol (HR 1.32, 1.05-1.66), and caused more adverse reactions (44% vs 33%). The trialists concluded the findings do not support it as a first-line treatment. Evidence: not-supported. (Source 3)
• Approved only as adjunctive therapy from age 12. The independent SANAD II generalised-epilepsy arm, which included generalised and unclassifiable epilepsy, found levetiracetam failed non-inferiority against valproate and was inferior per-protocol. Evidence: limited. (Source 4)
• Approved as adjunctive therapy only. In SANAD II's 520-patient generalised and unclassifiable epilepsy trial, levetiracetam did not meet non-inferiority against valproate for 12-month remission (HR 1.19, 95% CI 0.96-1.47) and had more adverse reactions (42% vs 37%). Evidence: limited. (Source 5)
• A 2025 systematic review and meta-analysis of 14 studies found no difference from phenytoin or fosphenytoin in seizure cessation (OR 1.18, 95% CI 0.94-1.48, p=0.16) but significantly fewer seizure recurrences (OR 0.60, 0.43-0.84) and fewer adverse events (OR 0.59, 0.37-0.94). It is as effective, not more effective. Evidence: established. (Source 6)
• Retrospective paediatric studies reported 41-66% improvement in behavioural adverse effects with pyridoxine, but a double-blind randomised placebo-controlled trial found improvement in both the pyridoxine and placebo groups over 8 weeks with no significant difference between them. Evidence: disputed. (Source 7)

Interactions

• Other antiepileptic drugs (carbamazepine, gabapentin, lamotrigine, phenobarbital, phenytoin, primidone, valproate) (pharmacokinetic study): Unusually for an antiseizure drug, levetiracetam neither raises nor lowers the blood levels of these drugs, and they do not change its levels. This was checked by measuring serum concentrations during the placebo-controlled trials. (Source 8)
• Pyridoxine (vitamin B6) (clinical trial): Pyridoxine is used off-label to try to blunt levetiracetam's behavioural side effects, not because it changes drug levels. The evidence is split, and the favourable figures are a 2026 review's second-hand summary of retrospective studies: those reported 41-66% improvement in behavioural effects and a lower discontinuation rate, but a double-blind randomised placebo-controlled trial found both the pyridoxine and placebo groups improved over 8 weeks with no significant difference between them. Pyridoxine was safe in that trial. (Source 7)
• Combined oral contraceptives (pharmacokinetic study): No interaction found in either direction. Levetiracetam 500 mg twice daily left the pharmacokinetics of an ethinyl estradiol plus levonorgestrel pill unchanged, along with luteinising hormone and progesterone levels, so loss of contraceptive protection is considered unlikely - which distinguishes it from enzyme-inducing antiseizure drugs. (Source 9)
• Warfarin (pharmacokinetic study): No interaction. Levetiracetam 1000 mg twice daily did not change the pharmacokinetics of R or S warfarin and did not affect prothrombin time; warfarin did not change levetiracetam levels. (Source 9)
• Digoxin (pharmacokinetic study): No interaction in either direction on drug levels or on the ECG, though note that digoxin appears elsewhere as a drug whose own safety depends on potassium and magnesium levels. (Source 9)
• Food (label): Food is not a barrier: levetiracetam is given orally with or without food, and the label puts no dietary restriction on it. Tablets must be swallowed whole rather than chewed or crushed. (Source 10)

Stopping it

• Levetiracetam must not be stopped abruptly. The label states it should generally be withdrawn gradually because of the risk of increased seizure frequency and status epilepticus, with rapid discontinuation reserved for a serious adverse reaction. (Source 11)
• Where the drug is stopped for a serious adverse reaction the label allows faster discontinuation, which is the trade-off the literature recognises between rebound seizure risk and harm from the drug. (Source 11)
• Behavioural and psychotic effects appear to reverse on stopping or reducing the dose. In the two adults hospitalised for psychosis in trials, both events developed within the first week of treatment and resolved within 1 to 2 weeks of stopping. (Source 12)
• Behavioural side effects are a common reason people come off it: 1.7% of adults in trials discontinued for behavioural adverse reactions versus 0.2% on placebo, and 11% of children had behavioural symptoms leading to discontinuation or dose reduction versus 6% on placebo. (Source 13)

What goes wrong

Levetiracetam causes non-psychotic behavioural symptoms in 13% of adults and 38% of children, against 6% and 19% on placebo. (Source 13)

• Randomized trial, High certainty.
• Size: pooled placebo-controlled clinical studies in adults and in children aged 4 to 16.
• Who: adults and children with epilepsy in placebo-controlled trials.
• How long: trial durations, typically 12 to 18 weeks.
• Result: Non-psychotic behavioural symptoms (aggression, agitation, anger, anxiety, apathy, depersonalisation, depression, emotional lability, hostility, hyperkinesias, irritability, nervousness, neurosis, personality disorder) in 13% of adults vs 6% of adult placebo, and 38% of children aged 4-16 vs 19% of child placebo. Irritability in 12% vs 0% in children aged 1 month to under 4 years. 1.7% of adults stopped for behavioural reactions vs 0.2% on placebo; 11% of children had behavioural symptoms leading to discontinuation or dose reduction vs 6% on placebo. A randomised double-blind placebo-controlled study found worsening on aggressive behaviour in children on the Achenbach Child Behavior Checklist.
• Funding: manufacturer registration programme, reported by the regulator.

In clinical studies, 13% of adult levetiracetam-treated patients and 38% of pediatric levetiracetam-treated patients (4 to 16 years of age) compared to 6% and 19% of adult and pediatric placebo-treated patients, experienced non-psychotic behavioral symptoms

Psychotic symptoms occurred in 1% of adults, 2% of children aged 4 to 16 and 17% of infants and toddlers on levetiracetam, versus 0.2%, 2% and 5% on placebo. (Source 12)

• Randomized trial, Moderate certainty.
• Size: pooled placebo-controlled studies across age groups.
• Who: adults and children with epilepsy.
• How long: trial durations; the two adult psychosis cases developed within the first week.
• Result: Psychotic symptoms 1% (adults) vs 0.2% placebo; 2% vs 2% in children 4-16; 17% vs 5% in children 1 month to under 4 years. Paranoia 1.6% vs 0%, confusional state 3.1% vs 0% in a controlled paediatric study. Two adults (0.3%) were hospitalised and stopped treatment for psychosis; both events developed in the first week and resolved within 1 to 2 weeks of stopping.
• Funding: manufacturer registration programme, reported by the regulator.

In clinical studies, 1% of levetiracetam-treated adult patients, 2% of levetiracetam-treated pediatric patients 4 to 16 years of age, and 17% of levetiracetam-treated pediatric patients 1 month to < 4 years of age experienced psychotic symptoms, compared to 0.2%, 2%, and 5% in the corresponding age groups treated with placebo.

Antiepileptic drugs as a class, levetiracetam included, roughly double the risk of suicidal thinking or behaviour, which in absolute terms is about one extra case per 530 people treated. (Source 14)

• Meta-analysis, Moderate certainty.
• Size: 199 placebo-controlled trials of 11 different AEDs; 27,863 drug-treated and 16,029 placebo-treated patients.
• Who: patients taking AEDs for epilepsy, psychiatric and other indications.
• How long: median treatment duration 12 weeks; most trials did not extend beyond 24 weeks.
• Result: Adjusted relative risk 1.8 (95% CI 1.2-2.7). Absolute incidence 0.43% on AEDs vs 0.24% on placebo, i.e. approximately one additional case of suicidal thinking or behaviour per 530 patients treated. For the epilepsy indication specifically, 3.4 per 1,000 on drug vs 1.0 per 1,000 on placebo, a risk difference of 2.4 per 1,000. Four suicides occurred in drug-treated patients and none on placebo, too few to draw a conclusion about completed suicide.
• Funding: regulator's pooled analysis of manufacturer trial data.

the estimated incidence rate of suicidal behavior or ideation among 27,863 AED-treated patients was 0.43%, compared to 0.24% among 16,029 placebo-treated patients, representing an increase of approximately one case of suicidal thinking or behavior for every 530 patients treated.

Pharmacovigilance and case data suggest a specific suicidality signal for levetiracetam, but the design cannot support a causal claim. (Source 15)

• Case series, Very low certainty.
• Size: pharmacovigilance reporting-odds-ratio analyses plus individual case reports.
• Who: people with epilepsy taking levetiracetam.
• How long: not applicable.
• Result: A reported odds ratio of 7.6 (95% CI 1.1-53.7) for suicidality in one analysis, with a very wide confidence interval; a separate FAERS analysis gave a reporting odds ratio of 1.286 (p<0.05). The review notes these are retrospective and rest on spontaneous adverse event reports with no adjustment for psychiatric comorbidity, and that people with epilepsy already have far higher baseline suicidality (pooled event rate 23.2% versus 3.1% in the general population).
• Funding: not stated.

Limit of this finding: The 7.6 figure is a reporting odds ratio taken by this review from another paper. It compares how often suicidality is mentioned in spontaneous adverse-event reports about levetiracetam against other drugs, so it measures reporting, not risk. Its range, 1.1 to 53.7, is so wide that the true value could be almost anything above 1. People with epilepsy also have much higher background rates of suicidality than the general population, and these reports are not adjusted for that. Read it as a signal worth watching, not as a measured seven-fold risk.

The use of LEV has been associated with an increased risk of suicidality in PWE, with a reported odds ratio (ROR) of 7.6 (95% confidence interval [CI] 1.1–53.7) ( 25 ). Since the introduction of LEV, several case reports have described the emergence or worsening of suicidality temporally associated with its use

Sleepiness and weakness are roughly twice as common on levetiracetam as on placebo, and worse at high doses given without titration. (Source 16)

• Randomized trial, High certainty.
• Size: controlled trials of adults with partial-onset seizures.
• Who: adults with epilepsy.
• How long: somnolence and asthenia occurred most frequently within the first 4 weeks.
• Result: Somnolence 15% on levetiracetam vs 8% on placebo, with no clear dose response up to 3000 mg/day but about 45% at 4000 mg/day given without titration; serious in 0.3% vs 0%; 3% stopped for somnolence vs 0.7% on placebo. Asthenia 15% vs 9%.
• Funding: manufacturer registration programme, reported by the regulator.

15% of levetiracetam-treated patients reported somnolence, compared to 8% of placebo-treated patients. There was no clear dose response up to 3000 mg/day. In a study where there was no titration, about 45% of patients receiving 4000 mg/day reported somnolence.

Levetiracetam can cause Stevens-Johnson syndrome and toxic epidermal necrolysis, typically 14 to 17 days after starting. (Source 17)

• Case series, Low certainty.
• Size: postmarketing reports in adults and children.
• Who: patients treated with levetiracetam.
• How long: median onset 14 to 17 days, but cases reported at least four months after starting.
• Result: Serious dermatological reactions including SJS and TEN reported in both paediatric and adult patients, with recurrence on rechallenge; the label says to stop at the first sign of rash unless clearly unrelated and never to resume if SJS/TEN is suspected.
• Funding: not applicable.

Serious dermatological reactions, including Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN), have been reported in both pediatric and adult patients treated with levetiracetam.

What the evidence supports

In the first pivotal placebo-controlled trial, levetiracetam added to existing treatment cut weekly partial seizure frequency by 26.1% to 30.1% over placebo. (Source 2)

• Randomized trial, Moderate certainty.
• Size: 293 randomised in Study 1 (placebo 95, 1000 mg/day 97, 3000 mg/day 101); 904 across the three pivotal trials.
• Who: adults with refractory partial-onset seizures for at least two years who had already taken two or more classical antiepileptic drugs, on a stable regimen of one or two AEDs.
• How long: 12-week baseline, then 18 weeks (6-week titration plus 12-week fixed dose)
• Result: Percent reduction in partial seizure frequency over placebo: 26.1% at 1000 mg/day and 30.1% at 3000 mg/day, both statistically significant versus placebo. Responder rate (50% or greater reduction) was a secondary outcome.
• Funding: manufacturer registration programme.

The primary measure of effectiveness was a between group comparison of the percent reduction in weekly partial seizure frequency relative to placebo over the entire randomized treatment period (titration + evaluation period). Secondary outcome variables included the responder rate (incidence of patients with ≥ 50% reduction from baseline in partial-onset seizure frequency). The results of the analysis of Study 1 are displayed in Table 10.

A 2026 review notes a Cochrane review concluded levetiracetam is effective and well tolerated for both focal and generalised epilepsies. (Source 18)

• Expert review, not systematic, Moderate certainty.
• Size: summary of a Cochrane review, not re-pooled.
• Who: people with focal and generalised epilepsies.
• How long: not stated.
• Result: Cochrane concluded levetiracetam is an effective and well-tolerated antiseizure medication for both focal and generalised epilepsies and an alternative to other NICE-recommended first-line drugs. This is a secondary report of a Cochrane conclusion rather than the review itself, and it sits alongside the independent SANAD II trials that found levetiracetam inferior as a first-line choice.
• Funding: not stated.

Limit of this finding: This is the 2026 review reporting what a Cochrane review concluded, not the Cochrane review itself, and nothing was re-pooled. It also sits against the independent SANAD II trials, which tested levetiracetam head to head over years and found it did not match lamotrigine or valproate as a first choice.

a recent Cochrane review concluded that LEV is an effective and well-tolerated ASM for the treatment of both focal and generalized epilepsies and identified it as an alternative to other first-line ASMs recommended by the National Institute for Health and Care Excellence in the United Kingdom

Levetiracetam has an unusually clean drug-interaction profile: it is not metabolised by liver cytochrome P450 enzymes and did not alter oral contraceptive, digoxin or warfarin handling. (Source 9)

• Blood level study, Moderate certainty.
• Size: dedicated clinical pharmacokinetic studies plus pharmacokinetic screening in the placebo-controlled epilepsy trials.
• Who: healthy volunteers and epilepsy trial participants.
• How long: short pharmacokinetic studies.
• Result: Levetiracetam 500 mg twice daily did not alter the pharmacokinetics of an oral contraceptive containing 0.03 mg ethinyl estradiol and 0.15 mg levonorgestrel, nor luteinising hormone or progesterone levels; 1000 mg twice daily did not alter digoxin pharmacokinetics or ECG effects, nor R and S warfarin pharmacokinetics or prothrombin time. It also does not alter concentrations of carbamazepine, gabapentin, lamotrigine, phenobarbital, phenytoin, primidone or valproate, and they do not alter its levels.
• Funding: manufacturer registration programme.

Levetiracetam (500 mg twice daily) did not influence the pharmacokinetics of an oral contraceptive containing 0.03 mg ethinyl estradiol and 0.15 mg levonorgestrel, or of the luteinizing hormone and progesterone levels, indicating that impairment of contraceptive efficacy is unlikely.

What the evidence does not support

Levetiracetam failed to prove non-inferior to lamotrigine as first-line treatment for newly diagnosed focal epilepsy and was worse on tolerability and cost-effectiveness. (Source 3)

• Randomized trial, High certainty.
• Size: 990 participants (lamotrigine 330, levetiracetam 332, zonisamide 328)
• Who: adults and children aged 5 years and older with newly diagnosed focal epilepsy and two or more unprovoked focal seizures, at UK neurology services.
• How long: randomised then followed a further 2 years.
• Result: Time to 12-month remission: levetiracetam versus lamotrigine HR 1.18 (97.5% CI 0.95-1.47), missing the non-inferiority limit of 1.329 which equates to an absolute difference of 10%. Per-protocol, lamotrigine was superior (HR 1.32, 97.5% CI 1.05-1.66). Adverse reactions in 144 (44%) on levetiracetam versus 108 (33%) on lamotrigine. Net health benefit 1.222 QALYs for levetiracetam versus 1.403 for lamotrigine.
• Funding: UK National Institute for Health Research Health Technology Assessment programme - independent of industry.

Levetiracetam did not meet the criteria for non-inferiority in the ITT analysis of time to 12-month remission versus lamotrigine (HR 1·18; 97·5% CI 0·95–1·47) but zonisamide did meet the criteria for non-inferiority in the ITT analysis versus lamotrigine (1·03; 0·83–1·28). The PP analysis showed that 12-month remission was superior with lamotrigine than both levetiracetam (HR 1·32 [97·5% CI 1·05 to 1·66]) and zonisamide (HR 1·37 [1·08–1·73]). There were 37 deaths during the trial. Adverse reactions were reported by 108 (33%) participants who started lamotrigine, 144 (44%) participants who started levetiracetam, and 146 (45%) participants who started zonisamide.

The SANAD II investigators concluded their findings do not support levetiracetam as a first-line treatment for focal epilepsy. (Source 19)

• Randomized trial, High certainty.
• Size: 990 participants.
• Who: as above.
• How long: minimum 2 years follow-up.
• Result: Interpretation: the findings do not support levetiracetam or zonisamide as first-line treatments and lamotrigine should remain first-line and the standard treatment in future trials.
• Funding: UK NIHR HTA programme.

These findings do not support the use of levetiracetam or zonisamide as first-line treatments for patients with focal epilepsy.

Levetiracetam also failed non-inferiority against valproate in newly diagnosed generalised and unclassifiable epilepsy, and was inferior per-protocol. (Source 20)

• Randomized trial, High certainty.
• Size: 520 participants (260 per arm)
• Who: people aged 5 years and older with newly diagnosed generalised (397) or unclassifiable (123) epilepsy; median age 13.9 years (range 5.0-94.4), 65% male.
• How long: randomised then followed a further 2 years.
• Result: Time to 12-month remission HR 1.19 (95% CI 0.96-1.47) against a non-inferiority margin of 1.314, so non-inferiority was not met; per-protocol, valproate was superior. Adverse reactions in 107 (42%) on levetiracetam versus 96 (37%) on valproate. Two deaths, one per group, unrelated to treatment. Levetiracetam was dominated by valproate on cost-utility. Valproate has its own well-known pregnancy risks that this comparison does not address.
• Funding: UK NIHR HTA programme - independent of industry.

Levetiracetam did not meet the criteria for non-inferiority in the ITT analysis of time to 12-month remission (HR 1·19 [95% CI 0·96–1·47]); non-inferiority margin 1·314. The PP analysis showed that the 12-month remission was superior with valproate than with levetiracetam. There were two deaths, one in each group, that were unrelated to trial treatments. Adverse reactions were reported by 96 (37%) participants randomly assigned to valproate and 107 (42%) participants randomly assigned to levetiracetam.

Adding pyridoxine (vitamin B6) to reduce levetiracetam's behavioural side effects looked promising in retrospective studies but failed in a placebo-controlled trial. (Source 7)

• Randomized trial, Low certainty.
• Size: a double-blind randomised placebo-controlled trial, set against retrospective paediatric series and one randomised case-controlled trial.
• Who: mostly children and adolescents with levetiracetam-related behavioural adverse effects.
• How long: 8 weeks in the placebo-controlled trial.
• Result: Retrospective paediatric data reported 41-66% improvement in behavioural adverse effects with pyridoxine and one randomised case-controlled trial found 92% of supplemented patients needed no dose adjustment. But the double-blind placebo-controlled trial showed improvement in both the pyridoxine and placebo groups over 8 weeks with no significant difference between groups at 8 weeks. Pyridoxine was safe. A systematic review found significant heterogeneity between studies.
• Funding: not stated.

Limit of this finding: The 41-66% improvement and the 92% figure are this review's summary of other researchers' retrospective studies, not new data, and the review describes one of them as 'a randomized case-controlled trial' - a label that contradicts itself, since a trial is either randomised or a case-control study and not both. The one finding here that comes from a properly blinded placebo-controlled trial is the negative one: both groups improved and there was no significant difference between them at 8 weeks.

In contrast, a double-blind randomized placebo-controlled trial showed improvement in BAEs in both the pyridoxine and placebo groups over an 8-week period, with no significant difference between groups at 8 weeks.

Where the evidence is mixed

In children with convulsive status epilepticus, levetiracetam stopped seizures no better than phenytoin or fosphenytoin but with fewer recurrences and fewer adverse events. (Source 6)

• Meta-analysis, Moderate certainty.
• Size: 14 studies pooled.
• Who: children with benzodiazepine-refractory convulsive status epilepticus.
• How long: acute treatment and in-hospital follow-up.
• Result: Seizure cessation OR 1.18 (95% CI 0.94-1.48, p=0.16) - no difference. Time to seizure termination mean difference -0.10 (95% CI -0.61 to 0.40, p=0.69) - no difference. Seizure recurrence OR 0.60 (95% CI 0.43-0.84, p=0.003) and adverse events OR 0.59 (95% CI 0.37-0.94, p=0.03), both favouring levetiracetam. No differences in mechanical ventilation, ICU admission or length of stay. The review notes the ConSEPT trial found levetiracetam not superior to phenytoin.
• Funding: not stated.

No significant difference was found between the two treatments regarding seizure cessation (odds ratio (OR): 1.18, 95% confidence interval (CI): 0.94–1.48; p = 0.16) or time to clinical seizure termination (mean difference: −0.10, 95% CI: −0.61 to 0.40; p = 0.69). However, levetiracetam was associated with significantly fewer seizure recurrences (OR: 0.60, 95% CI: 0.43–0.84; p = 0.003) and adverse events (OR: 0.59, 95% CI: 0.37–0.94; p = 0.03) compared with phenytoin/fosphenytoin.

Where the research disagrees

Whether levetiracetam should be a first-line antiseizure drug

• A 2026 narrative review summarising a Cochrane review, narrative review reporting a Cochrane conclusion, which itself pooled randomised trials: a recent Cochrane review concluded that LEV is an effective and well-tolerated ASM for the treatment of both focal and generalized epilepsies and identified it as an alternative to other first-line ASMs recommended by the National Institute for Health and Care Excellence in the United Kingdom (Source 18)
• SANAD II trialists (Marson and colleagues, 2021), NIHR-funded, open-label randomised non-inferiority trial in 990 patients with at least 2 years of follow-up and a cost-utility analysis: These findings do not support the use of levetiracetam or zonisamide as first-line treatments for patients with focal epilepsy. Lamotrigine should remain a first-line treatment for patients with focal epilepsy and should be the standard treatment in future trials. (Source 19)

Whether vitamin B6 (pyridoxine) reduces levetiracetam's behavioural side effects

• Retrospective paediatric studies and one randomised case-controlled trial, as summarised in a 2026 review, retrospective series and one randomised case-controlled trial: A retrospective pediatric study demonstrated that pyridoxine supplementation with LEV resulted in 41–66% improvement in BAEs ( 46 , 47 ). Another retrospective analysis showed that pyridoxine supplementation was associated with a lower rate of drug discontinuation compared with patients who did not receive supplementation ( 48 ). Marino et al. conducted a randomized case–controlled trial evaluating pyridoxine supplementation for LEV-related PBAEs in children and found that 92% of patients receiving pyridoxine supplementation did not require LEV dose adjustment or discontinuation. Improvement was observed after 9.06 ± 3.05 days of supplementation (Source 7)
• A double-blind randomised placebo-controlled trial, as summarised in the same review, double-blind randomised placebo-controlled trial over 8 weeks: a double-blind randomized placebo-controlled trial showed improvement in BAEs in both the pyridoxine and placebo groups over an 8-week period, with no significant difference between groups at 8 weeks (Source 7)

How much

• Reference intake: There is no reference intake for a drug; dose is set by the prescriber and depends on indication, age, formulation and kidney function. As a position, the 2026 US label starts adults 16 and over at 1000 mg/day for partial-onset seizures (500 mg twice daily), with increases of 1000 mg/day every 2 weeks. Paediatric dosing is by weight, for example 20 mg/kg/day rising to 60 mg/kg/day for ages 4 to under 16. (Source 21)
• Upper limit: No nutrient-style upper limit. As a position, the 2026 US label sets a maximum recommended daily dose of 3000 mg for partial-onset seizures in adults and states there is no evidence that doses above 3000 mg/day give any extra benefit. The label separately reports that about 45% of patients given 4000 mg/day without titration reported sleepiness. (Source 21)
• Studied: 1000 mg/day and 3000 mg/day versus placebo for 18 weeks in the first pivotal partial-onset seizure trial (293 patients). (Source 2)
• Studied: An initial advised maintenance dose of levetiracetam 500 mg twice daily for participants aged 12 and over in SANAD II, and 20 mg/kg twice daily for children aged 5 to 12. (Source 3)
• Studied: 500 mg twice daily in the oral contraceptive interaction study and 1000 mg twice daily in the digoxin and warfarin interaction studies. (Source 9)
• Studied: 4000 mg/day given without titration in a study where about 45% of patients reported somnolence. (Source 16)

A common belief, and what the research shows

The belief: Levetiracetam is the easy, safe, modern first-choice antiseizure drug - which is why it is prescribed so much.

What the research shows: Its popularity ran ahead of the evidence. A 2026 review reports it accounts for roughly 56% of US epilepsy prescriptions, but when independently tested head-to-head over years by the UK NIHR-funded SANAD II trials it failed non-inferiority twice: against lamotrigine in 990 people with focal epilepsy (HR 1.18, 97.5% CI 0.95-1.47, with 44% versus 33% adverse reactions) and against valproate in 520 people with generalised or unclassifiable epilepsy (HR 1.19, 95% CI 0.96-1.47). The trialists wrote that the findings do not support levetiracetam as a first-line treatment. Its drug-interaction profile really is unusually clean - but tolerability is not: behavioural symptoms hit 13% of adults and 38% of children against 6% and 19% on placebo.

Questions and answers

What is it?

Levetiracetam is a manufactured antiseizure medicine, not a nutrient. It is a single mirror-image form of a small pyrrolidone molecule, a white to off-white crystalline powder with a faint odour and a bitter taste, and it dissolves extremely well in water - about 104 g per 100 mL. It is chemically unrelated to the older antiepileptic drugs, which is part of why its side-effect and interaction profile is so different from theirs. (Source 10)

What does it do in the body?

It damps down the release of the brain's excitatory messenger chemicals. It binds a protein called SV2A on the synaptic vesicles - the small sacs nerve cells use to release their signals - which appears to reduce how readily glutamate is released. The label is honest that the molecular significance of that binding is not understood; the reason it is believed to be the mechanism is that, across a family of related molecules, how strongly they bind SV2A tracked how well they stopped seizures in mice. (Source 1)

Is it good or bad for you?

Genuinely useful for the right job, oversold for others. Added on in drug-resistant focal epilepsy it cut weekly seizures by 26% to 30% over placebo in the pivotal trials, and in children's status epilepticus it matches phenytoin with fewer recurrences and fewer side effects. But as a first-choice drug it lost twice in independent head-to-head trials, and its behavioural cost is real: behavioural symptoms in 13% of adults and 38% of children versus 6% and 19% on placebo, plus psychotic symptoms in 17% of the youngest children versus 5% on placebo. (Source 13)

How do you get more of it?

Levetiracetam is prescription-only in all its forms, so it is obtained only by prescription; no food or behaviour raises it. What the trials and the label record: adults 16 and over start at 1000 mg/day in two doses for partial-onset seizures and may go up by 1000 mg/day every 2 weeks to a maximum of 3000 mg/day, and the label states plainly that going above 3000 mg/day is not known to add benefit. It can be taken with or without food. This records trial and label doses and is not advice. (Source 21)

If it is harmful, what reduces it?

Where levetiracetam is causing harm, the literature says reduce or stop it - but gradually, because abrupt withdrawal risks more frequent seizures and status epilepticus. Behavioural and psychotic effects do appear to reverse: the two adults hospitalised for psychosis in trials both improved within 1 to 2 weeks of stopping. Adding vitamin B6 to blunt behavioural side effects looked good retrospectively but failed to beat placebo in a double-blind randomised trial. (Source 11)

Why might someone be low in it or missing it?

Nobody is deficient in levetiracetam - the body neither makes nor needs it. Blood levels do vary in ways that matter: it is cleared largely by the kidneys and is not extensively metabolised by liver cytochrome P450 enzymes, so kidney function drives dosing rather than liver enzyme interactions. Pregnancy lowers levels progressively, most markedly in the third trimester, so the label advises careful monitoring through pregnancy and the period after birth. (Source 8)

Which whole foods contain it or feed it?

None. Levetiracetam is a synthetic pyrrolidone made for tablets, solution and injection; no whole food contains it or feeds it. Food does not interfere with it either - it is given orally with or without food, unlike many other medicines whose absorption depends on what is in the stomach. Tablets must be swallowed whole rather than chewed or crushed. (Source 10)

What happens if you do not have it?

Not taking levetiracetam causes no deficiency; it suppresses seizures rather than supplying anything. What the placebo arms show is what happens without it in refractory focal epilepsy: people on levetiracetam had 26.1% to 30.1% fewer weekly partial seizures than those on placebo. For someone already established on it, stopping is a different question - abrupt withdrawal risks increased seizure frequency and status epilepticus. (Source 2)

How can you test for it?

There is no validated blood test used to guide levetiracetam treatment in ordinary care, and none appeared in the literature we searched; what gets measured is seizure frequency and kidney function, because dosing depends on renal clearance. Seizure diaries are the actual outcome measure - the pivotal trials used percent reduction in weekly partial seizure frequency and the responder rate, defined as a 50% or greater reduction from baseline. For the behavioural harms, validated questionnaires have been used in research, such as the Achenbach Child Behavior Checklist in a randomised paediatric study. (Source 13)

References

1. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 12.1 Mechanism of Action. 2026. Read the source
2. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 14.1 Partial-Onset Seizures (Study 1). 2026. Read the source
3. The Lancet. The SANAD II study of the effectiveness and cost-effectiveness of levetiracetam, zonisamide, or lamotrigine for newly diagnosed focal epilepsy: an open-label, non-inferiority, multicentre, phase 4, randomised controlled trial. 2021. PMID 33838757, DOI 10.1016/S0140-6736(21)00247-6. Read the source
4. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 1.2 Myoclonic Seizures in Patients with Juvenile Myoclonic Epilepsy. 2026. Read the source
5. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 1.3 Primary Generalized Tonic-Clonic Seizures. 2026. Read the source
6. Frontiers in Neurology. Levetiracetam versus phenytoin/fosphenytoin for second-line treatment of children with convulsive status epilepticus: an up-to-date meta-analysis and systematic review of randomized controlled trials. 2025. PMID 40510206, DOI 10.3389/fneur.2025.1580329. Read the source
7. Frontiers in Neurology. Overview of psychiatric and behavioral adverse effects of levetiracetam. 2026. PMID 42079817, DOI 10.3389/fneur.2026.1783559. Read the source
8. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 12.3 Drug Interactions (other antiepileptic drugs). 2026. Read the source
9. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 12.3 Drug Interactions (oral contraceptives, digoxin, warfarin). 2026. Read the source
10. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 11 DESCRIPTION. 2026. Read the source
11. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 5.8 Withdrawal Seizures. 2026. Read the source
12. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 5.1 Behavioral Abnormalities and Psychotic Symptoms (psychotic symptoms). 2026. Read the source
13. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 5.1 Behavioral Abnormalities and Psychotic Symptoms. 2026. Read the source
14. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 5.2 Suicidal Behavior and Ideation. 2026. Read the source
15. Frontiers in Neurology. Overview of psychiatric and behavioral adverse effects of levetiracetam. 2026. PMID 42079817, DOI 10.3389/fneur.2026.1783559. Read the source
16. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 5.3 Somnolence and Fatigue. 2026. Read the source
17. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 5.5 Serious Dermatological Reactions. 2026. Read the source
18. Frontiers in Neurology. Overview of psychiatric and behavioral adverse effects of levetiracetam. 2026. PMID 42079817, DOI 10.3389/fneur.2026.1783559. Read the source
19. The Lancet. The SANAD II study of the effectiveness and cost-effectiveness of levetiracetam, zonisamide, or lamotrigine for newly diagnosed focal epilepsy: an open-label, non-inferiority, multicentre, phase 4, randomised controlled trial. 2021. PMID 33838757, DOI 10.1016/S0140-6736(21)00247-6. Read the source
20. The Lancet. The SANAD II study of the effectiveness and cost-effectiveness of valproate versus levetiracetam for newly diagnosed generalised and unclassifiable epilepsy: an open-label, non-inferiority, multicentre, phase 4, randomised controlled trial. 2021. PMID 33838758, DOI 10.1016/S0140-6736(21)00246-4. Read the source
21. DailyMed (US National Library of Medicine), labeller Cardinal Health 107, LLC. LEVETIRACETAM tablet, film coated - Prescribing Information, 2.2 Dosing for Partial-Onset Seizures. 2026. Read the source