Migraine With Aura and Stroke After a Hospital Stay: What a 64.8 Million-Patient Study Found
In 64.8 million hospitalized US adults, migraine with aura was linked to about twice the rate of ischemic stroke readmission within 90 days. Migraine without aura showed no clear difference.
On this pageJump to a section7
What the Study Looked At
Migraine with aura has been linked to a higher long-term risk of ischemic stroke for years. Most of that research follows people for a decade or more. A new US study in Neurology asked a shorter-term question: when someone is admitted to the hospital and has migraine with aura on their record, how likely are they to come back with an ischemic stroke within the next 90 days?
The team, led by researchers at Brown University, used the Nationwide Readmissions Database, a national collection of US hospital records, for 2016 through 2019. This was a cohort study of every adult hospitalized during that period. Migraine was identified from diagnosis codes, either as the main reason for the stay or as a secondary diagnosis, and split into three groups: migraine with aura, migraine without aura, and migraine coded without a subtype. The comparison group was every adult admitted without any migraine diagnosis.
The outcome was a new hospital admission for acute ischemic stroke, the kind caused by a blocked artery, within 90 days.
Across the whole database, 1,257,671 adults (1.94%) had any migraine diagnosis during their stay, and 78,650 of them were admitted mainly for migraine. The average age of everyone admitted was 56, and 60% were women.
The Main Results
The researchers compared each migraine group with people admitted without migraine, after adjusting for cardiovascular risk factors. They used two methods to balance the groups: weighting based on each person's characteristics and Cox regression. Results are shown as hazard ratios, where 1.00 means no difference in the rate and 2.00 means roughly twice the rate. The ranges are 95% confidence intervals.
| Group (vs no migraine) | Adjusted hazard ratio for 90-day ischemic stroke | What it means |
|---|---|---|
| Migraine with aura | 1.98 (1.68 to 2.35) | About twice the rate of stroke readmission |
| Migraine without aura | 0.86 (0.62 to 1.21) | No clear difference |
| Migraine, subtype not coded | 0.96 (0.90 to 1.03) | No clear difference |
| Any migraine combined | 1.00 (0.94 to 1.06) | No difference |
Hazard ratios adjusted for cardiovascular risk factors. Intervals that cross 1.00 do not show a clear difference.
The signal was confined to migraine with aura. In the main analysis, people with migraine without aura or with migraine coded without a subtype did not have a higher rate of stroke readmission than people admitted without migraine. The authors note that modest associations for those groups appeared in some of their sensitivity analyses, so a small effect there cannot be ruled out.
This pattern matches earlier long-term research, where the extra stroke risk in migraine sits mostly with aura. We covered several of those studies recently, including a 1.7 million-woman Danish study of migraine, smoking and early stroke and a 2.2 million-person study of migraine, heart attack and stroke before and after menopause.
Reading the Numbers
A hazard ratio of 1.98 sounds large, but it is a relative figure. It describes how the rate in one group compares with another, not how many people had a stroke. Across all 64.8 million adults in the database, 272,812 people (0.42%) were readmitted with an ischemic stroke within 90 days. The paper's main result for the aura group is reported as a relative rate, and the absolute risk for an individual depends heavily on age and other conditions.
The population also matters. Everyone in this study was sick enough to be admitted to a hospital, for any reason, and the average age was 56. That is a very different group from a 30-year-old with occasional aura who has never been hospitalized. The authors frame their finding as applying to hospitalized patients.
This is an observational study built on billing codes. It shows that migraine with aura and early stroke readmission go together in hospital records. It cannot show that aura caused the strokes, and unmeasured factors may explain part of the link, a problem known as confounding.
What the Study Cannot Tell Us
Several features of the design limit how far the results go:
- Diagnosis codes, not clinical exams. Migraine, aura and stroke were all identified from hospital billing codes. The authors used standard, validated codes, but coding can still miss or mislabel cases, and some people with aura may have been coded as migraine without a subtype.
- Aura and stroke can look alike. In a hospital setting, symptoms recorded as aura may sometimes have been an early sign of a cerebrovascular problem. A database of codes cannot separate these cases.
- Only hospitalized adults. The results describe people admitted to the hospital, not people with migraine with aura in the general population.
- Unknown mechanism. The authors call for research into why migraine with aura is tied to this short-term stroke risk. The study does not include data on treatments, patent foramen ovale, or other possible explanations.
Patent foramen ovale, a small opening between the upper chambers of the heart, is one proposed link between aura and stroke. Our write-up of a 10-year study of heart shunts, migraine and stroke covers what is known about it.
What It Means for Patients and Clinicians
The authors conclude that the findings support careful vascular risk assessment in people with migraine with aura who are hospitalized. For most people with aura, the practical steps are the same ones that apply to long-term stroke risk.
- Make sure your clinician knows you have migraine with aura, not just migraine. The distinction changes some treatment choices, such as estrogen-containing birth control.
- Know what your usual aura looks like. If you are unsure which type you have, our guide to types of migraine aura describes them.
- Seek emergency care for sudden weakness, a drooping face, trouble speaking, an aura that is different from your usual pattern, or aura symptoms that last longer than an hour. Our article on migraine aura vs TIA covers how the two differ.
- Smoking, high blood pressure, high cholesterol and diabetes add to stroke risk and can be treated.
When a patient with migraine with aura is admitted for any reason, this study supports reviewing vascular risk factors before discharge and making sure follow-up covers them. It also supports recording the migraine subtype in the chart, since the association in this study was limited to aura. Our overview of primary stroke prevention in migraine summarizes the modifiable factors.
Key Takeaways
- Among 64.8 million US adults admitted to the hospital from 2016 to 2019, those with migraine with aura had about twice the rate of readmission for ischemic stroke within 90 days (adjusted hazard ratio 1.98).
- Migraine without aura and migraine coded without a subtype showed no clear difference in the main analysis.
- The study used hospital billing codes and included only hospitalized adults, so it shows an association and does not prove cause.
- The findings support checking vascular risk factors in people with migraine with aura, especially around a hospital stay.
This article is for educational purposes only and does not replace professional medical advice, diagnosis, or treatment. Talk with a qualified clinician about your own stroke risk and treatment options. If you have sudden weakness, numbness, trouble speaking, vision loss, or the worst headache of your life, call emergency services right away.
References
- Strelecky L, Shu L, Spears R, et al. Ninety-Day Acute Ischemic Stroke Readmission in People Hospitalized With Migraine With Aura: A National Cohort Study. Neurology. 2026;107(12):e218771. doi:10.1212/WNL.0000000000218771
Get studies like this by email
Every Friday, the week's migraine studies and news explained in plain language. No ads.


This article is for education and is not medical advice. Talk with your own clinician before starting, stopping or changing any treatment.


