Migraine is one of the few areas where supplements have been tested seriously enough for a professional body to name specific compounds at specific doses. That makes it unusual, and it makes the honest version of this article awkward for us — because the compounds with the best evidence are mostly ones we don't sell, and the one we do sell is at the wrong dose in the wrong form.
Here is what the guidelines actually say, graded by how strong the evidence is. If it's magnesium specifically you're here for, we've gone through the clinical evidence for magnesium and migraine on its own terms — this piece is the wider comparison.
The evidence tiers
The clearest grading comes from the joint American Academy of Neurology and American Headache Society guideline on complementary treatments for episodic migraine prevention. It sorted them by evidence strength, and the ranking surprises most people:
| Level | Meaning | Treatment and dose |
|---|---|---|
| A | Established as effective | Petasites (butterbur) 50–75 mg twice daily |
| B | Probably effective | Riboflavin 400 mg/day · Magnesium 600 mg trimagnesium dicitrate daily · Feverfew · histamine (subcutaneous) |
| C | Possibly effective | CoQ10 100 mg three times daily |
The American Headache Society's own summary of nutraceuticals for migraine prevention names three — magnesium at 400–500 mg daily as magnesium oxide, riboflavin at 400 mg daily, and CoQ10 at 300 mg daily — with the rationale being magnesium's effect on cortical excitability and glutamate, riboflavin's role in mitochondrial energy production, and CoQ10's observed reduction in attack frequency.
Note the discrepancy on magnesium, because it's instructive rather than a problem: one source specifies 600 mg of a citrate, the other 400–500 mg as the oxide. Different trials used different salts at different doses. What both agree on is that the dose is high and the form is specified — which is exactly what most magnesium marketing omits.
Butterbur is rated highest, and we'd still be careful
The top-graded option is the one almost nobody recommends any more, and the reason is worth knowing.
Butterbur plants naturally contain pyrrolizidine alkaloids, which are toxic to the liver and carcinogenic. Commercial preparations are supposed to have them removed, but cases of liver injury associated with butterbur products led to withdrawals and regulatory restrictions in several markets. A Level A rating for efficacy does not carry a Level A rating for safety, and the two are separate questions.
We mention it because leaving out the highest-rated option would misrepresent the evidence, and because it illustrates something about this category: "it works" and "you should take it" are not the same claim. If you are considering butterbur, it is a conversation with a doctor and a pharmacist about a specific purified product, not a shelf decision.
The doses are much higher than typical supplement doses
400 mg of riboflavin is roughly 300 times the daily reference intake. That isn't a rounding error — it's a different use of the molecule. At that dose it is being used for a mitochondrial effect, not to correct a deficiency.
Which means a general B-complex will not do this. A typical multivitamin contains a couple of milligrams. If you are taking riboflavin for migraine specifically, the dose is the entire point, and a product that doesn't state it in hundreds of milligrams is not the intervention that was studied.
The same logic applies to CoQ10 at 300 mg daily and magnesium in the 400–600 mg range. These are therapeutic doses, and they are well above what general wellness products contain.
Our magnesium is not the studied intervention
Proco's magnesium glycinate delivers 275 mg of elemental magnesium. The migraine research used 400–500 mg as the oxide, or 600 mg as a citrate. Ours is below either dose, in a third form that neither guideline names.
We chose glycinate deliberately, for gut tolerability at a nightly dose and for sleep, which is what the product is for — and the timing question follows from that. It is not a migraine prevention product and we're not going to relabel it as one because migraine is a bigger search term.
If you want the studied intervention, you want a different product from a different company, and you want to read the next section before you buy it.
We don't sell the other two at all
No riboflavin, no CoQ10, no feverfew, no butterbur. If the evidence supports them at those doses — and professional bodies say it does, at varying strengths — then the honest thing is to say so and point you elsewhere, rather than steer you to the thing we happen to have.
What to know if you go looking
Riboflavin turns urine bright yellow. That's harmless and expected, and worth knowing so it doesn't alarm you. It is water-soluble and generally well tolerated even at 400 mg, which is part of why it's an attractive option.
CoQ10 comes as ubiquinone or ubiquinol; absorption differs and so does price. Take it with a fatty meal — it's fat-soluble, and taken on an empty stomach much of it is wasted. Two interactions worth raising with a pharmacist: it may reduce the effect of warfarin, and statins lower the body's own CoQ10 levels, which is a common reason people are taking it in the first place.
Magnesium at 400–600 mg can loosen the bowels, and this is strongly form-dependent — oxide is the worst offender and is also one of the forms that was studied. That trade-off is real, and titrating up slowly helps. Magnesium also needs care in significant kidney impairment, where it can accumulate.
Feverfew should not be stopped abruptly after prolonged use, and it is avoided in pregnancy.
All of them take time. These are preventives, not painkillers. Trials typically run two to three months before judging the result. If you try one for a fortnight and stop, you haven't tested it — you've just spent the money.
And none of them are for pregnancy without advice, which matters here because migraine is common in women of reproductive age and several standard prescription preventives are also contraindicated.
If you're already taking prescription medication, particularly anything psychiatric, check the combination before adding any of this — we've set out the interactions that apply to our own range.
Before any of this: is it actually migraine?
The dose question is downstream of the diagnosis question, and this is where a lot of money gets wasted. Tension-type headache does not respond to migraine-specific treatment, cluster headache needs entirely different and highly effective treatments, and headache on most days in someone taking frequent painkillers may be being caused by the painkillers. The criteria that separate them are here, and getting the category right changes everything that follows.
One more worth excluding: a headache present on waking that eases through the morning, alongside unrefreshing sleep, can point to obstructive sleep apnea rather than a primary headache disorder.
What this doesn't replace
Prescription preventives and the newer CGRP-targeting treatments are a different order of effect from anything on this page. If your migraines are frequent or disabling, a supplement is not the conversation to be having instead of that one — it's the conversation to have alongside it, with a doctor who knows what else you're taking.
Doses and evidence levels as published by the American Academy of Neurology and American Headache Society. This article is not medical advice, and high-dose supplementation should be discussed with a doctor or pharmacist, particularly alongside prescription medication or in pregnancy.
Sources: AAN/AHS evidence-based guideline: NSAIDs and other complementary treatments for episodic migraine prevention in adults · American Headache Society — nutraceuticals for migraine prevention