Here is something I tell every epilepsy patient, and something most neurologists mention once in passing and never follow up on: your seizure medication is almost certainly depleting nutrients that matter.
Anti-epileptic drugs (AEDs) — now increasingly called anti-seizure medications (ASMs) — are essential. If you have epilepsy, you need them. Nothing in this article should be interpreted as a reason to reduce or stop your medication. But the metabolic cost of long-term AED use is real, it is well-documented, and it is dramatically under-addressed in routine clinical care.
The Problem: Drug-Induced Nutrient Depletion
Most AEDs have been in use for decades, and the literature on their metabolic effects is mature. The mechanisms are well understood: some AEDs induce hepatic cytochrome P450 enzymes, which accelerate the metabolism of fat-soluble vitamins. Others interfere with folate metabolism. Others deplete specific cofactors through mechanisms that are still being elucidated.
The result is a predictable pattern of deficiencies that varies by medication class. And these are not trivial deficiencies — they contribute to osteoporosis, anemia, neuropathy, fatigue, and in some cases, worsening seizure control.
I check vitamin D levels in every patient on an enzyme-inducing AED, and the deficiency rate is staggering. In my practice, roughly 70–80% of patients on carbamazepine or phenytoin long-term have insufficient or deficient vitamin D levels. This is not an anomaly — it is consistent with the published literature.
Enzyme-Inducing AEDs: Carbamazepine, Phenytoin, Phenobarbital
The classic enzyme-inducing AEDs are the biggest offenders. Carbamazepine (Tegretol), phenytoin (Dilantin), and phenobarbital all strongly induce CYP450 enzymes in the liver, which accelerates the breakdown of vitamin D and other fat-soluble vitamins.
Vitamin D Depletion
This is the most clinically significant depletion. Enzyme-inducing AEDs increase catabolism of 25-hydroxyvitamin D, the storage form measured by standard blood tests. The result is progressive vitamin D deficiency, which in turn impairs calcium absorption and leads to decreased bone mineral density.
The data on this is unambiguous. Multiple studies have shown that patients on long-term enzyme-inducing AEDs have a 2–6 fold increased risk of osteoporotic fractures compared to age-matched controls. For a neurological condition that already carries an inherent fall risk from seizures, this is a serious compounding problem.
What I recommend: Vitamin D3 (cholecalciferol) at 2,000–4,000 IU daily, titrated to maintain a serum 25(OH)D level of 40–60 ng/mL. I check levels every 6 months in patients on enzyme-inducing AEDs. Some patients require 5,000 IU or more. Standard multivitamin doses of 400–800 IU are woefully insufficient for this population.
Calcium
Because vitamin D depletion impairs calcium absorption, supplemental calcium is often necessary — but not always. I prefer to optimize vitamin D first and recheck calcium and parathyroid hormone levels before adding calcium supplementation. Over-supplementation of calcium carries its own cardiovascular risks.
When I do supplement: Calcium citrate 500–600mg daily (citrate is better absorbed than carbonate, especially in patients with reduced stomach acid). Always taken separately from AED doses by at least 2 hours, as calcium can reduce absorption of some medications.
Folate
Phenytoin in particular interferes with folate absorption and metabolism. Chronic folate deficiency can cause megaloblastic anemia, fatigue, and cognitive changes that are often mistakenly attributed to the seizure disorder itself.
What I recommend: Folic acid 1mg daily for all patients on phenytoin. For women of childbearing age on any AED, folate supplementation is critical — the AAN specifically recommends this due to the elevated risk of neural tube defects. I prescribe 1–4mg daily depending on pregnancy planning status, in close coordination with their obstetrician.
Valproate (Depakote)
Valproate has its own distinct depletion profile. It does not induce CYP450 enzymes — in fact, it is an enzyme inhibitor. But it causes nutrient depletion through different mechanisms.
L-Carnitine
This is the depletion that keeps me up at night. Valproate directly inhibits the carnitine transport system and increases renal excretion of carnitine. Carnitine deficiency impairs mitochondrial fatty acid oxidation, which can cause hepatotoxicity, hyperammonemia, and in severe cases, fatal hepatic failure.
The risk is highest in children, in patients on polytherapy, and in patients with underlying mitochondrial disorders (which are more common in the epilepsy population than most clinicians appreciate). But even adult patients on valproate monotherapy can develop subclinical carnitine deficiency that manifests as fatigue, muscle weakness, and cognitive sluggishness.
What I recommend: L-carnitine supplementation at 990mg twice daily (the typical over-the-counter dosing) for all patients on valproate. For patients with documented carnitine deficiency or symptoms suggestive of depletion, I use prescription levocarnitine (Carnitor) at higher doses. This is not optional — this is standard-of-care supplementation that should be discussed at initiation of valproate therapy.
Folate (Again)
Valproate is a known folate antagonist. This is especially critical for women of childbearing age, given valproate's well-documented teratogenicity. The combination of valproate's direct teratogenic effects and its folate-depleting properties creates a compounding risk for neural tube defects.
My practice: I have an extended conversation with every woman of childbearing age before prescribing valproate. If we proceed with valproate, high-dose folate (4mg daily) is non-negotiable. In many cases, I am transitioning women to alternative AEDs entirely — but that is a separate clinical discussion.
Levetiracetam (Keppra): The "Newer" Drug That Is Not Immune
Levetiracetam has been called "metabolically clean" because it does not induce or inhibit hepatic enzymes. This has led to a widespread assumption that it does not cause nutrient depletion. That assumption is wrong — or at minimum, incomplete.
Vitamin D
Emerging evidence suggests that levetiracetam is associated with vitamin D deficiency, though the mechanism is less clear than with enzyme-inducing AEDs. A 2019 study by Harijan et al. found that children on levetiracetam monotherapy had significantly lower vitamin D levels than controls. Adult data is more mixed, but the signal is consistent enough that I now check vitamin D in all my epilepsy patients, regardless of which AED they are on.
What I recommend: Vitamin D3 at 1,000–2,000 IU daily as a baseline for patients on levetiracetam, with dose adjustment based on lab levels.
Pyridoxine (Vitamin B6)
There is emerging — though not yet definitive — evidence that levetiracetam may affect pyridoxine (B6) metabolism. Some clinicians, including myself, have observed that B6 supplementation appears to mitigate the irritability and behavioral side effects ("Keppra rage") that are the most common reason patients discontinue the drug. This is anecdotal and mechanistically plausible but not yet supported by randomized controlled trials.
My practice: For patients experiencing irritability on levetiracetam, I trial pyridoxine 50mg daily before adding a second psychotropic medication. It helps some patients, it is cheap, and at that dose, it is safe. I am transparent with patients that this is off-label and based on clinical observation, not Level A evidence.
Why "Just Take a Multivitamin" Fails
I hear this constantly — from patients, from other physicians, from well-meaning pharmacists. The reasoning sounds logical: if AEDs deplete multiple nutrients, why not cover everything with a multivitamin?
The answer is dose. A standard multivitamin contains 400 IU of vitamin D, 200mg of calcium, 400mcg of folic acid, and no L-carnitine. For a patient on carbamazepine who needs 4,000 IU of vitamin D, 500mg of calcium citrate, and 1mg of folic acid, a multivitamin covers approximately 10–40% of what they actually need.
A multivitamin is not a therapeutic intervention. It is a safety net designed for generally healthy people eating an imperfect diet. AED-induced nutrient depletion requires targeted, medication-specific supplementation at therapeutic doses. A multivitamin gives you a false sense of coverage while leaving the actual deficiencies unaddressed.
Building Your Protocol
Here is how I approach supplementation for my epilepsy patients:
- Identify the AED class and its known depletion pattern
- Check baseline labs — at minimum: 25(OH)D, calcium, CBC (for folate-related anemia), and carnitine if on valproate
- Supplement at therapeutic doses specific to the identified depletions
- Recheck labs every 6–12 months and adjust doses accordingly
- Coordinate with the prescribing neurologist — never add supplements without informing the physician managing your seizure medications
The depletion patterns are predictable and the supplementation protocols are straightforward. The barrier is not complexity — it is awareness. If your neurologist has not discussed nutrient depletion with you, bring this article to your next appointment and start the conversation.