What the Science Says
Plenty of people with migraine will tell you they know when the weather is about to turn. The conviction is widespread, it is sincere, and it is worth taking seriously.
It is also, at the population level, not well supported. That combination is the interesting part, and it points somewhere more useful than the usual article on this subject does.
What the research actually shows
A systematic review published in 2025 gathered 14 studies covering 2,696 participants and found the results inconsistent. Several studies reported associations between pressure drops or rapid fluctuations and increased migraine frequency. Fewer found any association with severity. None found a link to attack duration. The reviewers noted that most of the included studies relied on self-reported diaries without standardised severity measures, used small samples over short periods, measured pressure in different ways, and mostly failed to adjust for other weather variables or for individual differences in sensitivity.
A prospective study in the Boston area followed 98 adults with episodic migraine through electronic daily diaries, accumulating 4,406 days of observation with matched weather and air pollution data. It found generally null associations between barometric pressure and migraine onset, and noted that this agreed with the majority of earlier studies. In the same body of work, an analysis that looked at participants individually found the direction of the association was inconsistent even among people identified as weather-sensitive.
So the honest summary: there may be a real effect on how often attacks occur, it is weak and inconsistent in aggregate, and the evidence does not support confident claims about thresholds or mechanisms.
Anyone telling you that a drop of a specific number of hectopascals over a specific number of hours is your trigger has gone considerably beyond what the literature can support.
Why that doesn’t mean you’re imagining it
Two things can be true at once, and here they probably are.
The recurring criticism in this literature is precisely that studies fail to account for individual variability in weather sensitivity. If pressure genuinely affects a minority of people with migraine and does nothing to the majority, averaging across a cohort washes the signal out. A weak population average is exactly the shape you’d expect from a real effect in a subgroup.
Which reframes the useful question. Not “does barometric pressure trigger migraine,” which the field has been arguing about for fifty years without resolution, but “does it trigger mine.” That is answerable, by you, and it is the reason a personal log is worth keeping here even though the general literature is a mess.
The mechanisms, with appropriate hedging
Migraine biology is reasonably well understood in outline. Activation of the trigeminovascular system, release of neuropeptides including CGRP, and neurogenic inflammation with sensitisation of pain pathways. This is solid, and it is why CGRP-targeted drugs work.
What is not established is the route from a falling barometer to that cascade. The proposed pathways are plausible and largely untested in humans:
- Pressure differentials across air-filled cavities. Middle ear and paranasal sinuses equalise imperfectly during rapid change, and the resulting mechanical stress could stimulate trigeminal endings. Plausible; not demonstrated as a migraine trigger.
- Inner ear effects. Pressure changes could alter the mechanics of the vestibular apparatus, producing conflicting spatial signals in a nervous system already prone to over-responding. This is the most intuitively appealing route for vestibular symptoms specifically, and it remains hypothesis.
One correction worth making, because it appears in a great deal of writing on this topic. The story where serotonin spikes, cerebral vessels constrict, serotonin then plunges and vessels rebound into dilation to produce the throbbing pain is the old vascular theory of migraine. It has been substantially superseded. Migraine is now understood as a primarily neurological disorder in which vascular changes are a consequence rather than the cause. Articles still reciting the constrict-then-dilate sequence are working from a model the field moved on from.
The attribution problem, which is bigger than it looks
This is the part that changes how you should track.
Perceived triggers rarely survive testing. In a study of 328 people with episodic migraine who logged daily for 90 days, participants endorsed an average of 28 potential triggers each. When each person’s data was tested individually, an average of 2.2 triggers per person were statistically associated with increased attack risk. People believe in roughly thirteen times more triggers than hold up.
Some “triggers” are the attack starting. The premonitory phase begins hours before the headache and produces yawning, neck tension, food cravings, fatigue, difficulty concentrating, and heightened sensitivity to light and sound. If you become sensitive to light two hours before the pain and then blame bright light, you have inverted cause and effect. Research directly comparing reported triggers with premonitory symptoms in the same patients supports this reading for several commonly cited triggers.
Weather is unusually exposed to this problem. If the premonitory phase makes you feel heavy, foggy and vaguely unwell, and you look outside at a darkening sky, the attribution writes itself. The attack was already underway.
And you can’t blind the weather. You always know what the sky is doing, which makes expectation impossible to remove from self-report.
None of this means weather does nothing. It means the belief that it does is not evidence, and that separating the two requires a specific kind of record.
Vestibular migraine is a diagnosis, not a description
Worth being precise, because “vertigo plus headache” is not the same as vestibular migraine, and the distinction matters clinically.
The Bárány Society and International Headache Society criteria require at least five episodes of vestibular symptoms of moderate or severe intensity lasting between 5 minutes and 72 hours, a current or previous history of migraine, and migraine features accompanying at least half the episodes.
The differential matters. Vestibular migraine is the second most common cause of episodic vertigo, after benign paroxysmal positional vertigo. Others include Ménière’s disease, which overlaps enough to be a genuine diagnostic challenge in early years, vestibular neuritis, persistent postural-perceptual dizziness, superior canal dehiscence, episodic ataxia, and transient ischaemic attack. These have different treatments and different prognoses.
Seek prompt medical assessment for vertigo accompanied by any new neurological signs — weakness, numbness, slurred speech, double vision, difficulty walking — for sudden hearing loss, or for a first severe episode with no prior history. Sudden vertigo can be a stroke, and that is not something to log and review in a fortnight.
If your vertigo is positional and lasts seconds, that pattern points elsewhere and is often very treatable with a repositioning manoeuvre.
Treatment, which the tracking does not replace
The previous version of this article offered hydration, steady sleep and limiting sensory strain. Those are reasonable, and they are nowhere near the whole picture.
Migraine is one of the more treatable neurological conditions. Acute options and preventive options both exist, including the CGRP-targeted therapies, and effective prevention typically reduces attack frequency far more than trigger avoidance does.
For vestibular migraine specifically, honesty is required: treatment is largely empirical and lacks high-quality trial evidence. Approaches in use include vestibular rehabilitation, standard oral migraine preventives, and more recently CGRP-targeted treatments, with antihistamines and benzodiazepines sometimes used for acute relief. The evidence for triptans in vestibular migraine remains uncertain.
That uncertainty is a reason to see a neurologist or a specialist dizziness clinic, not a reason to substitute an app for one.
How to track this properly
If you want an answer specific to you, the method matters more than the diligence.
Pre-specify. Decide before you start that you are testing pressure change against attack onset. Do not log twelve variables and go looking. With enough variables, something always correlates.
Log before you check the weather. Rate your state first, look at conditions afterwards. Otherwise you are measuring your expectations.
Log premonitory symptoms separately. Yawning, neck tension, food cravings, fogginess, unusual fatigue. If these consistently precede the weather you’re blaming, that is your answer, and it is a genuinely valuable one: it means you have hours of warning and can treat early, which is when acute migraine treatment works best.
Record the time of onset, not just the day. Pressure moves over hours. Daily resolution loses the effect you’re looking for.
Give it months, not weeks. You need enough attacks and enough pressure events that don’t coincide. Two weeks cannot do this.
Expect a negative result. Given that roughly two of twenty-eight believed triggers survive testing, the most likely outcome is that pressure isn’t yours. That would be worth knowing, because it frees you from arranging your life around the forecast.
Where SharedSky fits
SharedSky records pressure alongside your entries, which is what makes the above possible without a spreadsheet. Our separate piece on building a baseline covers the method in more detail.
What the app can honestly offer: a time-stamped record with environmental context attached, kept in a form you can hand to a neurologist. Given that migraine diagnosis and management both depend heavily on attack history, and that most people reconstruct that history badly from memory in a ten-minute appointment, this is the real payoff. It is more valuable than any correlation you might find.
What it cannot offer: a verdict. If your log suggests a pressure pattern, that is a hypothesis worth mentioning to your doctor, not a diagnosis.
SharedSky is not a medical device and does not provide medical advice. Recurrent vertigo warrants proper diagnosis. Seek urgent care for vertigo with new neurological symptoms or sudden hearing loss. See our Medical Disclaimer.
Sources
- Impact of Barometric Pressure Changes on the Severity, Frequency, and Duration of Migraine Attacks: A Systematic Review of the Literature, Cureus (2025). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12617017/
- Weather, ambient air pollution, and risk of migraine headache onset among patients with migraine, Environment International (2019). https://www.sciencedirect.com/science/article/pii/S0160412019300662
- An observational study of self-reported migraine triggers and prospective evaluation of the relationships with occurrence of attacks enabled by a smartphone application, Headache (2022). https://pubmed.ncbi.nlm.nih.gov/35670125/
- Karsan et al., Are some patient-perceived migraine triggers simply early manifestations of the attack?, Journal of Neurology (2021). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8068686/
- Schulte et al., Photo-, osmo- and phonophobia in the premonitory phase of migraine: mistaking symptoms for triggers?, Journal of Headache and Pain (2015). https://pubmed.ncbi.nlm.nih.gov/25904144/
- Lempert et al., Vestibular migraine: diagnostic criteria, consensus document of the Bárány Society and the International Headache Society. https://pubmed.ncbi.nlm.nih.gov/34719447/
- Insights into Vestibular Migraine: Diagnostic Challenges, Differential Spectrum and Therapeutic Horizons (2025). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12294935/
- Editorial: Vestibular migraine, Frontiers in Neurology (2026). https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2025.1587097/full
Ready to uncover your weather sensitivity thresholds? Start logging your daily baseline with SharedSky today.