Beyond the Pill

Your brain weighs about three pounds. It consumes 20% of all the energy your body produces. And it has no backup generator.

That last fact matters if you have migraine. Because a growing body of evidence — from neuroimaging labs in Germany, clinical trials in Switzerland, meta-analyses pooling thousands of patients — points to a single, startling idea: your migraine may be, at its root, an energy crisis.

Magnetic resonance spectroscopy has shown that between attacks, the brains of people with migraine produce up to 16% less ATP — the molecule that powers every cell — than the brains of people without.^[1] The neurons still fire. The thoughts still form. But the margin is thinner. The system is more fragile. And when demand spikes — from stress, weather, hormones, a skipped meal — there isn't enough fuel to keep the grid stable.

The lights flicker. Then they go out.

This is called the mitochondrial hypothesis of migraine. And it's the reason three ordinary substances — a mineral, a B vitamin, and a coenzyme — have become some of the most studied supplements in headache medicine. Not because they kill pain. Because they feed the power plants.

But a plausible theory is not proof. What matters is whether real patients in controlled trials got fewer migraines. Let's look at what they found.

Magnesium: The Deficit You Can't See on a Lab Report

Start with a number: up to 50% of people with migraine have low magnesium levels.^[2] That statistic comes from studies measuring ionized magnesium during acute attacks — work pioneered in the early 1990s by Dr. Alexander Mauskop at the New York Headache Center, who has spent more than three decades studying this single mineral in migraine patients.^[3]

Now here's the problem. Your doctor orders a blood test. It comes back normal. You both move on.

But that normal result may be meaningless. Only about 2% of your body's magnesium circulates in the blood. The rest sits inside cells and bones, invisible to routine labs. The American Migraine Foundation has stated plainly that standard blood tests "do not accurately measure magnesium levels in the brain."^[4] You can be deficient where it counts and look fine on paper.

So the clinical question becomes simpler: does giving people magnesium reduce their migraines — whether or not we can prove they were low?

The answer, across multiple lines of evidence, is yes.

A 2016 meta-analysis published in Pain Physician pooled data from 21 randomized controlled trials. Oral magnesium significantly reduced both the frequency and intensity of migraine attacks.^[5] A systematic review of placebo-controlled trials from 1990 to 2016 found reductions in migraine days ranging from 22% to 43%.^[6] And in 2012, the American Academy of Neurology and the American Headache Society — the two organizations that set the standard for migraine treatment in the United States — reviewed this evidence and assigned magnesium a Level B rating: probably effective, should be considered for prevention.^[7]

The strongest results appear in two groups: people who experience aura, and women whose migraines cluster around menstruation.^[4] If either describes you, this isn't a "maybe worth trying" conversation. It's a clinical one.

Which form, and how much?

Here's where the details matter — and where most advice falls short.

Magnesium oxide is the most commonly studied form. It's cheap. It doesn't need a prescription. The American Migraine Foundation recommends it at 400–600 mg per day.^[4] But its bioavailability is roughly 4% — meaning 96% of what you swallow never reaches your cells.^[8] And at high doses, it reliably causes diarrhea. One trial actually failed to show a benefit, likely because half the treatment group dropped out due to GI side effects from a poorly absorbed magnesium salt.^[9]

Magnesium citrate absorbs far better and was the form used in the key positive trial showing significant reductions in migraine frequency at 600 mg per day.^{[6, 8]} Magnesium glycinate is gentler still — well-absorbed, easy on the stomach, and increasingly the form headache specialists recommend for long-term daily use.^[10] Magnesium threonate can cross the blood-brain barrier, which makes it theoretically appealing, but migraine-specific data is still thin.^[10]

The practical move: start at 200 mg of citrate or glycinate. Build to 400–600 mg over two weeks, split into morning and evening. Give it a full three months. If one form bothers your stomach, switch — don't quit.

Riboflavin: Two Dollars a Month, 59% Responder Rate

Vitamin B2 does not have the charisma of magnesium. No one has built a wellness brand around it. It turns your urine bright yellow and has all the mystique of a fluorescent office light.

It may also be the most underrated tool in migraine prevention.

The trial that changed the field was published in Neurology in 1998. It was small — 55 patients — but it was randomized, double-blinded, and placebo-controlled, the gold standard of clinical design. Patients took either 400 mg of riboflavin or a placebo every day for three months.^[11]

The results: 59% of those on riboflavin experienced at least a 50% reduction in attacks. On placebo, just 15%. The number-needed-to-treat was 2.3 — which means that for every 2.3 people who take riboflavin, one gets a clinically meaningful benefit.^[11] Compare that to topiramate, a frontline prescription preventive, which has a number-needed-to-treat of about 3.5 to 6, depending on the trial — along with side effects that include cognitive fog, weight loss, tingling, and kidney stones.

Was the riboflavin trial too small to trust on its own? Fair question. That's why the 2021 meta-analysis matters. Researchers pooled nine controlled trials — a combined 673 patients — and confirmed: 400 mg per day of riboflavin for three months significantly reduced migraine days, frequency, duration, and pain scores.^[12] A comprehensive 2024 expert review in Headache concluded that riboflavin "can be recommended for migraine prevention in adults" with minimal adverse events.^[13]

The mechanism is elegant and specific. Riboflavin is the raw material for FAD and FMN, two coenzymes that sit at the heart of the mitochondrial electron transport chain.^[14] Without them, energy production stutters. In a brain that's already operating on narrow margins, that stutter can trigger the cascade — cortical spreading depression, inflammation, pain.

What you need to know before starting: The dose is 400 mg per day. Not 25 mg (the amount some multivitamins contain). Not 100 mg (the dose in two pediatric trials that failed to show a benefit — likely because the dose was too low).^[15] You need 400 mg, taken consistently, for at least three months. Benefits typically emerge around week eight. Anyone who stops after a month stopped the experiment before it began.

Side effects: bright yellow urine. That's it.

CoQ10: Real Evidence, Realistic Expectations

Coenzyme Q10 completes the mitochondrial trio. It's a fat-soluble antioxidant that your body produces naturally and that plays a direct role in ATP synthesis — the same energy pathway that magnesium and riboflavin support.^[16]

The key randomized trial came from the University of Zurich. Forty-two migraine patients received either 300 mg of CoQ10 or placebo for three months. Among those on CoQ10, 48% saw their attack frequency cut in half. On placebo: 14%.^[17] A separate trial found that after three months, CoQ10 performed as well as amitriptyline — a widely prescribed migraine preventive — but with fewer side effects.^[18]

The most comprehensive test of these findings came in 2021: a meta-analysis of six randomized controlled trials totaling 371 participants. CoQ10 reduced migraine frequency by about 1.5 attacks per month and shortened headache duration. It did not, however, significantly reduce pain severity.^[19]

The AAN/AHS classified CoQ10 as Level C: possibly effective.^[7] That's one tier below magnesium and two below the prescription medications at Level A. It means the data is promising but not definitive — the trials are small, and the effects are more modest.

Still, for patients who want to avoid prescription side effects or the risk of medication overuse headache, or who want to layer a supplement on top of existing treatment, CoQ10 has a defensible role. The studied dose is 100–300 mg daily, taken with food to improve absorption.^[20] It is well-tolerated. It does not interact with most medications. And it costs less than a coffee.

Three Supplements the Evidence Does Not Support

Honesty about what doesn't work is as important as clarity about what does.

Butterbur once held the highest evidence rating of any migraine supplement — Level A, established as effective.^[7] Then, in 2015, the AAN retracted the recommendation.^[21] The plant contains pyrrolizidine alkaloids — chemicals that damage the liver, and that independent testing found in one-third of commercial products.^[22] The American Headache Society now cautions against its use.^[20] A supplement that works but may quietly injure your liver is not a solution.

Feverfew has been studied for decades. A 2020 Cochrane review of six trials found a small effect — roughly 0.6 fewer attacks per month than placebo.^[23] The 2024 expert review noted mixed results, inconsistent dosing across studies, and insufficient safety data, concluding that caution is warranted.^[13]

Melatonin (typically 3 mg at bedtime) has very low certainty of evidence for migraine prevention specifically. The same 2024 review added a practical warning for the U.S. market: purity of commercially sold melatonin products is "very poor."^[13] If you have sleep problems and migraine, it may help indirectly. As a dedicated migraine preventive, the evidence isn't there.

Five Questions That Turn a Supplement Into a Strategy

A bottle on your nightstand is not a plan. A conversation with your doctor is. Here are the questions worth bringing:

"Could I have a magnesium deficit that standard labs would miss?" This is especially likely if you have aura, menstrual migraine, diabetes, or take diuretics.^[4]

"What form and dose should I start with?" Glycinate for a sensitive stomach. Citrate if constipation is an issue. Oxide if cost is the priority. 400–600 mg daily.^{[4, 8]}

"Can I add riboflavin to what I'm already taking?" It's exceptionally safe, but your doctor should see the full picture. One study suggested riboflavin and beta-blockers may work through different mechanisms, potentially complementing each other.^[14]

"How long before I should expect results?" Three months. Minimum. The studies are unanimous on this.^{[11, 12, 17]}

"Should I track my attacks?" Yes. A simple daily log — headache or no headache, severity from 1 to 10 — is the only honest way to separate real improvement from the hope that something is working.

What This Adds Up To

No supplement will cure migraine. Anyone who tells you otherwise is selling something.

But the evidence for these three — magnesium, riboflavin, and CoQ10 — is not a marketing claim. It comes from randomized controlled trials published in peer-reviewed journals. It's been evaluated by the American Academy of Neurology and the American Headache Society. It's grounded in a biological mechanism — mitochondrial energy metabolism — that researchers have been building, testing, and refining for over two decades.

Magnesium: Level B evidence. Migraine days reduced by 22–43% across multiple trials.^{[5, 6]} Riboflavin: 59% responder rate in the landmark trial, confirmed by meta-analysis of 673 patients.^{[11, 12]} CoQ10: Level C evidence, approximately 1.5 fewer attacks per month, with a safety profile most prescriptions can't match.^{[17, 19]}

They cost less combined than a single copay. They require no prescription. And their side effects — diarrhea from magnesium if you choose the wrong form, yellow urine from riboflavin — are the kind most migraine patients would gladly trade for what they endure now.

These supplements won't silence the pain on the day a migraine arrives. But they may slowly, quietly reduce how many of those days there are. Not by masking symptoms. By helping your brain hold its ground.

One steadier morning at a time.