This post covers how vestibular migraine is actually diagnosed, which popular diet and supplement advice stands up, and where testing earns its place and where it doesn't.
How I grade evidence in this post: strong trial evidence · mixed (trials disagree, or one trial only) · mechanistic and untested · association only · none. Expert consensus is labelled as consensus, because it isn't trial evidence.
Why it gets missed
Each symptom gets looked at on its own. The dizziness goes to one specialist, the fog to another, and the nausea gets put down to something else. Nobody joins them up. That's the sum-of-the-parts problem this practice is built around, and vestibular migraine is one of the clearest examples of it.
It isn't only patients who struggle. In a 2025 survey of Canadian ear, nose and throat surgeons, only 15.9% said they'd had extensive training in vestibular migraine during their specialist training (Chowdhury et al., 2026). Treat that as a signal, not a measurement: only 44 surgeons replied, a 7.4% response rate.
How common is it? The international diagnostic criteria put it at between 1% and 2.7% of the general population, with 65–85% of patients female (Lempert et al., 2022). I haven't found a UK figure.
How it's actually diagnosed
Grade: international consensus criteria, not trial evidence.
The criteria were set jointly by the Bárány Society and the International Headache Society in 2012, and left unchanged in the 2022 update (Lempert et al., 2022). In plain English, you need all four:
- At least five episodes of vertigo, or of dizziness brought on by moving your head with nausea. They must be moderate (getting in the way of daily life) or severe (stopping it), and last between 5 minutes and 72 hours.
- A history of migraine, now or in the past.
- A migraine feature in at least half of the episodes. That can be a migraine-type headache, or sensitivity to both light and sound, or a visual aura (flickering zigzag lights that spread over minutes). One feature is enough.
- Nothing else explains it better.
So the headache-free part is real, but it needs stating carefully. The dizzy episodes themselves don't have to come with a headache. What the criteria do require is a migraine history at some point, which may have been years ago. Without that history, the most a specialist can call it is probable vestibular migraine.
There's no blood test, scan or ear test that confirms it. The diagnosis rests entirely on what you report. Ear and balance tests can be abnormal during or just after an episode, but between episodes they usually aren't specific enough to help. That's one reason a careful history matters so much, and why it gets missed.
What has to be ruled out matters.
- Benign paroxysmal positional vertigo (BPPV): brief spinning set off by turning over in bed or looking up. Vestibular migraine can mimic it.
- Ménière's disease: attacks with hearing loss in one ear, tinnitus and ear fullness. The two can overlap, and early Ménière's can look like vestibular migraine.
- Problems with blood flow to the back of the brain, particularly in older people with vascular risk factors.
- Anxiety-related dizziness. More than half of people with vestibular migraine also have an anxiety or mood disorder, which can complicate the picture in both directions.
Sorting these out is a doctor's job, and often a specialist's.
Call 999 if dizziness or vertigo comes on suddenly with any of these: a drooping face, weakness or numbness in an arm or leg, slurred or muddled speech, double vision or loss of vision, difficulty walking or standing, or a sudden severe headache unlike anything you've had before. Sudden neck pain with vertigo also needs urgent assessment. A stroke can look like "just vertigo" (NHS stroke symptoms).
Sudden hearing loss in one or both ears also needs same-day help: contact NHS 111 or your GP the same day.
What a trigger diary shows
Grade: none for triggers as a diagnostic test; a diary is still the most useful thing you can bring.
The criteria authors say this plainly: menstruation, stress, lack of sleep, dehydration and some foods can all trigger attacks, but none of them is a diagnostic criterion, because how well they predict vestibular migraine hasn't been properly studied (Lempert et al., 2022).
That's not a reason to skip the diary. It's a reason to treat it as your own evidence rather than a checklist borrowed from somebody else. The diagnosis rests on timing, episode length and what comes with each attack, and a diary records exactly that.
Record every episode against sleep, the menstrual cycle, meals and missed meals, alcohol, caffeine, screens and busy visual environments, stress, and weather if you suspect it. We use a 28-day grid for this. Four weeks covers a full cycle and enough episodes to see a pattern rather than guess at one.
Diet and supplement claims, graded
Popular articles on vestibular migraine get the diagnostic story right, then switch to guessing when they reach treatment. Here's what the evidence supports.
Something worth knowing first: prescription drugs aren't better evidenced here either. The criteria authors note that evidence for treating vestibular migraine with anti-migraine drugs is insufficient, and mostly observational rather than from randomised trials (Lempert et al., 2022). Every claim below, drug or supplement, borrows its evidence from headache migraine. I apply the same standard to both.
"Cut seed oils and eat more oily fish"
Grade: mixed — one good trial, in headache migraine, with the benefit on a secondary outcome.
This is the strongest dietary evidence available, and popular articles tend not to cite it. A 16-week randomised trial of 182 adults with frequent migraine (88% women) compared three diets (Ramsden et al., 2021):
- Omega-3 up: EPA plus DHA raised to 1.5 g a day.
- Omega-3 up and linoleic acid down: as above, with linoleic acid (the main fat in seed oils) cut to 1.8% of energy or less.
- Control: a typical diet.
Cutting linoleic acid alongside the omega-3 increase reduced headache days by about 4 a month compared with control, and by about 2 a month more than raising omega-3 alone.
The limits, stated plainly:
- One of the trial's two primary outcomes, a headache quality-of-life score, did not improve significantly. Headache days were a secondary outcome.
- It was headache migraine. Nobody has run this trial in vestibular migraine.
- The omega-3 increase did much of the work.
For scale: 1.8% of energy is about 4 g of linoleic acid a day on a 2,000 kcal diet. That's my arithmetic, not the trial's. Popular advice to keep it under 5 g a day lands near the trial's level. Advice to aim for 2 g goes beyond anything the trial tested.
The claim that linoleic acid mimics oestrogen, and so explains why more women are affected, has no trial behind it. Grade: mechanistic and untested.
"Take magnesium — glycinate or malate"
Grade: mixed for magnesium in headache migraine. The forms usually recommended have not been tested.
The magnesium evidence in migraine comes from specific forms and doses, and the trials disagree:
| Trial | Form and dose | Result |
|---|---|---|
| Peikert et al., 1996, n=81, 12 weeks | Trimagnesium dicitrate, 600 mg (24 mmol) a day | Attacks fell 41.6%, against 15.8% on placebo. Diarrhoea in 18.6% |
| Pfaffenrath et al., 1996, n=69, stopped early | Magnesium aspartate, 10 mmol twice a day | No benefit: 28.6% responded on magnesium, 29.4% on placebo |
| Karimi et al., 2021, n=63 analysed, crossover | Magnesium oxide, 500 mg (the paper doesn't say whether that is per tablet or per day, or the weight of the compound or of the magnesium in it) | Similar to the drug valproate. No placebo group, so it can't show magnesium beats placebo |
A 2012 American neurology and headache guideline rated magnesium "probably effective" for preventing migraine (Level B), alongside riboflavin, and rated coenzyme Q10 "possibly effective" (Level C) (Holland et al., 2012). It's based on trials published up to 2009.
Glycinate and malate, the forms usually recommended online, haven't been tested for migraine prevention. A PubMed search for migraine trials of either form found none. They're chosen for tolerability, which is a reasonable reason, but not the same thing as evidence for the indication. My own magnesium page said otherwise until October, and has been corrected.
For dizziness and balance specifically, the evidence is thinner still. A 2026 review found only early-phase trials and case series, and called any use of magnesium in ear and balance disorders beyond hearing-loss prevention experimental (Büntzel et al., 2026). The authors belong to the German Society of Magnesium Research.
"Riboflavin (vitamin B2) for an energy-starved brain"
Grade: mixed — one small positive trial in headache migraine; untested in vestibular migraine.
In a three-month randomised trial of 55 people with migraine, 400 mg of riboflavin a day reduced attack frequency and headache days compared with placebo. 59% halved their headache days, against 15% on placebo (Schoenen et al., 1998). That's promising but small. The "energy-starved brain" explanation is still a hypothesis: mechanistic and untested as a reason it works.
Combination products
Grade: mixed. A three-month trial of a product combining magnesium, riboflavin and coenzyme Q10 (sold as Dolovent in the UK; one of the trial's authors works for its sponsor) did not significantly reduce migraine days: 6.2 to 4.4 a month, against 6.2 to 5.2 on placebo. It did reduce pain intensity and a headache-burden score (Gaul et al., 2015). People may genuinely feel better on it. That isn't the same as fewer attacks.
"Low vitamin D is behind it"
Grade: association only. In one hospital's records, people with vestibular migraine had lower vitamin D than people with other causes of dizziness. But vitamin D didn't make the final diagnostic model (Zhang et al., 2025), and a low level isn't proof that taking vitamin D prevents attacks.
The point that ties these together
A nutrient taken at many times the dietary amount, indefinitely, is being used as a drug. That isn't wrong. But it deserves a drug's questions: why, for how long, and how will we know if it's working? More in Topping Up a Shortfall, or Taking a Drug?
And the other side of it: if you're genuinely short of something, that shortfall is the bottleneck, and correcting it can make a real difference. If you aren't, more of it is a pile at the wrong station. That's the argument of Why More of a Thing Usually Does Nothing, and it's why the next section exists.
One source commonly listed for these popular claims is an online forum thread, which is a good reason to read If You Can't Check It, It Isn't Evidence.
Where testing earns its place
This is the Test, Don't Guess part. Testing doesn't diagnose vestibular migraine. Nothing does except the clinical history. What testing does is replace assumptions with measurements before anyone starts supplementing.
| Question | What can answer it | Grade |
|---|---|---|
| Do attacks follow the hormone cycle? | The 28-day diary first; a DUTCH hormone test timed to the cycle alongside it | Diary: your own data. Using DUTCH to guide migraine treatment: mechanistic and untested |
| Is vitamin D actually low? | Blood test, not assumption | Measuring is standard. Supplementing to prevent attacks: untested |
| Is magnesium actually low? | Blood test, knowing its limit: almost all the body's magnesium sits in bone and tissue, not blood, so a normal serum result doesn't rule out a shortfall | Measure, and know the limit |
| Is the "energy-starved brain" idea relevant to you? | Organic acids (OAT) markers linked to riboflavin and energy metabolism | Mechanistic and untested for guiding migraine treatment |
Declared interest: these tests are part of what I offer. None of them has been shown to improve outcomes in vestibular migraine, and I won't tell you they have. What they do is stop a supplement being taken for a shortfall nobody measured.
Normal imaging and normal ear tests are expected in vestibular migraine, and they aren't a reason to stop looking. See Your Scan Is Probably Normal For Your Age for why "normal" is both less reassuring and less alarming than it sounds.
What to do
- See your GP first, and ask directly whether vestibular migraine has been considered. Mention any migraine history, even years ago. Take your diary.
- Keep the 28-day diary before changing anything.
- If you want to try diet, the best-evidenced change is more oily fish and fewer seed oils, from a trial in headache migraine. Give it three to four months, not three weeks.
- If you want to try a supplement, choose a form and dose that has actually been tested. Agree a time limit, and judge it against your diary, not your hopes.
- Test what can be tested before supplementing for a shortfall you haven't measured. Blood testing options · The TDG programme
What we don't know
- No diet or supplement trial has been run in vestibular migraine specifically. Neither, mostly, have drug trials.
- Whether hormone or organic acids testing changes treatment decisions for the better in migraine is untested.
- How common vestibular migraine is in the UK. The 1–2.7% figure is international.
- Which triggers genuinely predict attacks. The criteria authors say it hasn't been properly studied.
References
- Büntzel J, et al. The role of magnesium in otology – a narrative review. Magnes Res 2026;38(4):126–135. PMID 42439044. doi:10.1684/mrh.2026.0553
- Chowdhury R, et al. Current practices in vestibular migraine management among Canadian otolaryngologists: a national survey. Audiol Res 2026;16(3). PMID 42345622. doi:10.3390/audiolres16030082
- Gaul C, et al. Improvement of migraine symptoms with a proprietary supplement containing riboflavin, magnesium and Q10. J Headache Pain 2015;16:516. PMID 25916335. doi:10.1186/s10194-015-0516-6
- Holland S, et al. Evidence-based guideline update: NSAIDs and other complementary treatments for episodic migraine prevention in adults. Neurology 2012;78(17):1346–1353. PMID 22529203. doi:10.1212/WNL.0b013e3182535d0c
- Karimi N, et al. The efficacy of magnesium oxide and sodium valproate in prevention of migraine headache: a randomized, controlled, double-blind, crossover study. Acta Neurol Belg 2021;121(1):167–173. PMID 30798472. doi:10.1007/s13760-019-01101-x
- Lempert T, et al. Vestibular migraine: diagnostic criteria (update). J Vestib Res 2022;32(1):1–6. PMID 34719447. doi:10.3233/VES-201644 (full text read)
- NHS 111 Wales. Stroke: symptoms. 111.wales.nhs.uk/Stroke. Accessed 3 Oct 2026.
- Peikert A, et al. Prophylaxis of migraine with oral magnesium. Cephalalgia 1996;16(4):257–263. PMID 8792038. doi:10.1046/j.1468-2982.1996.1604257.x
- Pfaffenrath V, et al. Magnesium in the prophylaxis of migraine — a double-blind placebo-controlled study. Cephalalgia 1996;16(6):436–440. PMID 8902254. doi:10.1046/j.1468-2982.1996.1606436.x
- Ramsden CE, et al. Dietary alteration of n-3 and n-6 fatty acids for headache reduction in adults with migraine: randomized controlled trial. BMJ 2021;374:n1448. PMID 34526307. doi:10.1136/bmj.n1448
- Schoenen J, et al. Effectiveness of high-dose riboflavin in migraine prophylaxis. Neurology 1998;50(2):466–470. PMID 9484373. doi:10.1212/wnl.50.2.466
- Zhang S, et al. A multimodal data-based diagnostic model for predicting vestibular migraine: a retrospective study. Front Neurol 2025;16:1723008. PMID 41414993. doi:10.3389/fneur.2025.1723008