GLP-1 With Keppra, Lamictal, and Other Antiseizure Medications

Roughly 3.4 million US adults live with active epilepsy (Zack 2017^[1]), and obesity prevalence in that population tracks the general population — about 40%. The clinical question for prescribers is whether a GLP-1 receptor agonist can be layered onto an existing antiseizure medication (ASM) regimen safely, and whether the choice of ASM matters for weight outcomes. Two reviews answer the weight half (Ben-Menachem 2007^[2], Biton 2003^[3]): ASMs split into weight-gaining (valproate, gabapentin, pregabalin, carbamazepine), weight-neutral (levetiracetam, lamotrigine, lacosamide, brivaracetam, eslicarbazepine), and weight-losing (topiramate, zonisamide) buckets. The pharmacokinetic interaction half is shorter: GLP-1 peptides do not go through CYP450, so the classic enzyme-induction and protein-binding interactions that dominate ASM polypharmacy do not apply (Patsalos 2020^[7], Patsalos 2008^[6]). The remaining question is seizure threshold during GLP-1 dose escalation, and the published psychiatric-safety post hoc from the STEP program (Wadden 2024^[9]) reported no seizure signal.

The honest summary

• ASM weight effects matter more than the GLP-1 interaction. Valproate (Depakote) can add 5–10 kg and partially offset a GLP-1; switching to a weight-neutral ASM where the indication allows is often higher-yield than adding any obesity medication (Ben-Menachem 2007^[2], Biton 2003^[3]).
• GLP-1 peptides do not use CYP450. Semaglutide, tirzepatide, and liraglutide are degraded by ubiquitous peptidases, not hepatic enzymes. Carbamazepine, phenytoin, and other strong CYP3A4 inducers do not reduce GLP-1 exposure in any clinically meaningful way (mechanism reviewed in Patsalos 2020^[7]).
• Keppra (levetiracetam) and Lamictal (lamotrigine) are the easiest stacks. Both are weight-neutral (Ben-Menachem 2007^[2]) and neither shares a clearance pathway with GLP-1 agonists (renal for levetiracetam; UGT1A4 glucuronidation for lamotrigine).
• Topiramate is the unusual case. It is an effective ASM AND an FDA-approved weight-loss agent (as phentermine/topiramate ER, Qsymia, pivotal CONQUER trial Gadde 2011^[5]). Stacking topiramate + a GLP-1 for a patient with epilepsy AND obesity is biologically synergistic but requires attention to bicarbonate, nephrolithiasis risk, and cognitive side effects.
• No published seizure signal with GLP-1. The Wadden 2024 STEP psychiatric-safety post hoc^[9] reviewed STEP 1, 2, 3, and 5 (~3,300 participants) and found no convulsion or seizure imbalance vs placebo.

How ASMs sort by weight effect

Ben-Menachem 2007^[2] and Biton 2003^[3] remain the canonical reviews of ASM body-weight effects. The pattern is well-replicated:

• Weight gain (avoid or switch where possible): valproate (Depakote) +5 to +10 kg over 6–12 months; gabapentin (Neurontin) +5 kg; pregabalin (Lyrica) +3 kg; carbamazepine (Tegretol) modest gain.
• Weight neutral (preferred for obese epilepsy patients): levetiracetam (Keppra), lamotrigine (Lamictal), lacosamide (Vimpat), brivaracetam (Briviact), eslicarbazepine (Aptiom), perampanel (Fycompa), phenytoin (Dilantin).
• Weight loss: topiramate (Topamax) about −5 kg over 6 months at typical ASM doses; zonisamide (Zonegran) about −3 kg (Gadde 2003 RCT in obese non-epileptic adults^[4]).

The clinical hierarchy for an obese epilepsy patient: if seizure control allows, switch valproate to a weight-neutral or weight-losing ASM before adding obesity pharmacotherapy. If seizure control demands valproate, a GLP-1 still works — it just has more weight to fight against.

Pharmacokinetic interactions: short list

GLP-1 receptor agonists are large peptides degraded by ubiquitous peptidases (DPP-4, neprilysin) rather than hepatic cytochromes. The dominant ASM interaction mechanisms — CYP3A4 induction by carbamazepine, phenytoin, and phenobarbital, and UGT competition between valproate and lamotrigine — do not touch GLP-1 exposure (Patsalos 2020^[7]). The remaining considerations are downstream of GLP-1 mechanism rather than direct pharmacokinetics:

• Gastric delay and oral ASM absorption. Tirzepatide and semaglutide slow gastric emptying modestly during titration. For oral ASMs with narrow therapeutic windows (phenytoin, carbamazepine, valproate), this can shift the absorption curve. Keep dosing time consistent and consider a level at week 12 of GLP-1 titration if seizure control wavers (Patsalos 2008 ILAE TDM guidelines^[6]).
• Lamictal and tirzepatide. Lamotrigine is cleared by UGT1A4. Tirzepatide gastric delay can shift peak absorption but does not alter overall exposure in published modeling; no dose change is required.
• Valproate and the pancreatitis overlap. Valproate carries a small but real pancreatitis risk; GLP-1 agonists carry a small acute-pancreatitis label warning. The combined risk is not additive in the published data, but baseline lipase and a low threshold for re-checking on persistent abdominal pain is the conservative practice.
• Eslicarbazepine and hyponatremia. Eslicarbazepine and oxcarbazepine can cause SIADH-pattern hyponatremia. GLP-1 dose escalation drives fluid loss (nausea, reduced intake, occasional vomiting). Baseline sodium plus a recheck at week 4 catches most cases.

Seizure threshold during GLP-1 escalation

The published evidence does not show a seizure signal with GLP-1 receptor agonists. The STEP psychiatric-safety post hoc (Wadden 2024 JAMA Internal Medicine^[9]) pooled STEP 1, 2, 3, and 5 (~3,300 semaglutide participants vs placebo) and reported no convulsion or seizure imbalance. Manufacturer pharmacovigilance reports through 2024 have not surfaced a class-level seizure signal for tirzepatide or liraglutide either.

Two indirect triggers do exist and deserve flagging at the patient-counseling stage:

• Hypoglycemia in patients on insulin or sulfonylureas. GLP-1 agonists in combination with insulin or sulfonylureas raise hypoglycemia risk, and severe hypoglycemia is a well-documented seizure trigger. Patients with epilepsy who are also on basal insulin need an insulin dose reduction (typically 20–30%) at GLP-1 initiation.
• Dehydration during nausea peaks. The first 4–8 weeks of titration produce the highest GI side-effect burden. Significant volume depletion plus electrolyte derangement (low sodium, low magnesium) can lower seizure threshold. Standing electrolyte counseling (sodium, potassium, magnesium) during titration is reasonable for any patient with epilepsy.

Drugs to avoid stacking on a GLP-1 in epilepsy

Three medications commonly considered for weight loss or depression in obesity carry seizure-threshold issues that make them poor stacking partners when the patient has active epilepsy:

• Bupropion (Wellbutrin). The product label carries a contraindication for patients with a seizure disorder. Avoid bupropion in active epilepsy even if obesity pharmacotherapy is desired; the Wellbutrin/Contrave route is closed. See our GLP-1 + Wellbutrin stacking review for the broader safety picture.
• Tramadol. Tramadol lowers seizure threshold even in patients without epilepsy and is the most common pharmacological cause of seizures encountered in emergency departments outside of overdose contexts. Avoid in active epilepsy regardless of GLP-1 status.
• Stimulants (phentermine, methylphenidate, amphetamines). Stimulants are generally avoided in poorly controlled epilepsy. Stable, well- controlled epilepsy patients may use stimulants under neurology supervision, but the combination is higher-risk than a clean GLP-1 monotherapy.

Magnitude: weight change at 12 months by ASM and GLP-1 combination

Ketogenic diet, epilepsy, and the theoretical GLP-1 synergy

The classical ketogenic diet (4:1 fat-to-protein-plus-carb) and the modified Atkins diet (MAD) are evidence-based non-pharmacological options for drug-resistant epilepsy in adults (Cervenka 2016, the Johns Hopkins Adult Epilepsy Diet Center series^[8]). The same diets reliably produce weight loss through carbohydrate restriction and appetite suppression via circulating ketone bodies. The theoretical case for combining a GLP-1 with a keto/MAD in a patient with refractory epilepsy AND obesity is strong: the diet addresses seizure burden while the GLP-1 accelerates weight loss; ketones and GLP-1 both suppress appetite, which may make the diet easier to maintain.

No randomized trial has tested this combination. The practical caveats are familiar to keto clinicians: monitor urinary stones (topiramate + keto raises stone risk further), serum bicarbonate, and lipid panel; expect constipation; and plan for the muscle-mass protection protocol covered in our GLP-1 muscle-loss prevention review.

Practical protocol for an epilepsy patient starting a GLP-1

1. Baseline labs. CBC, comprehensive metabolic panel including sodium and bicarbonate, lipase, and an ASM level for narrow-window agents (valproate, phenytoin, carbamazepine) before GLP-1 initiation (Patsalos 2008 ILAE TDM^[6]).
2. Switch the offending ASM if feasible. For patients on valproate or gabapentin with well-controlled seizures, discuss switching to levetiracetam, lamotrigine, or lacosamide with the treating neurologist before adding obesity pharmacotherapy.
3. Standard GLP-1 titration with neurology awareness. The 4-week step-up schedule for semaglutide or tirzepatide does not need modification for stable epilepsy. Communicate the new medication to the neurology team.
4. Hydration and electrolyte counseling. Standing instructions for sodium, potassium, and magnesium intake during the nausea-prone titration window. For eslicarbazepine or oxcarbazepine patients, repeat serum sodium at week 4.
5. ASM level at week 12. Recheck the relevant ASM level once the GLP-1 maintenance dose has stabilized to confirm absorption has not shifted (valproate, phenytoin, carbamazepine).
6. Insulin and sulfonylurea reduction at start. If the patient is on insulin or a sulfonylurea, reduce the dose by 20–30% at GLP-1 initiation to limit hypoglycemia-triggered seizures.
7. Pregnancy planning. ASM teratogenicity planning is a neurology decision (valproate is contraindicated; lamotrigine and levetiracetam are preferred). GLP-1 agonists should be washed out at least 8 weeks before attempting pregnancy per the semaglutide and tirzepatide labels.

What good provider follow-up looks like

For most epilepsy + obesity patients, the right configuration is co-management: the prescribing neurology team owns the ASM choice and seizure-control endpoint; the obesity medicine provider owns GLP-1 titration and weight endpoint. Shared documentation of the ASM choice, baseline labs, and the planned 12-week recheck prevents the common failure mode of one team adjusting a medication without the other knowing.

Insurance access is generally favorable: ASMs are broadly covered as Schedule IV or non-controlled medications, and obesity-indicated GLP-1 coverage now depends on BMI plus comorbidity wording, which active epilepsy does not block.

Related research and tools

• GLP-1 brain fog and cognitive effects — the cognitive-side-effect literature relevant to patients already on topiramate or zonisamide
• GLP-1 with lithium and mood stabilizers — companion review for patients with overlapping neuropsychiatric medications
• GLP-1 rechallenge after pancreatitis — relevant when valproate is the historical pancreatitis trigger
• GLP-1 + Wellbutrin stacking — why bupropion is contraindicated in active epilepsy
• GLP-1 muscle-loss prevention protocol — protein and resistance-training protocol that carries over for keto + epilepsy patients

Important disclaimer. This article is educational and does not constitute medical advice. ASM changes are neurology decisions and must not be made unilaterally by obesity medicine providers or patients. Bupropion is contraindicated in patients with seizure disorder. Pregnancy planning for women on ASMs and GLP-1 agonists requires coordinated neurology and obstetric input. PMIDs were verified live against the PubMed E-utilities API on 2026-05-29.

Last verified: 2026-05-29. Next review: every 12 months, or sooner if new pharmacovigilance data on GLP-1 seizure signal or new ASM weight-effect randomized data is published.