A Genetic Deep Dive to Find What’s Rare in Migraine

By Neil Andrews | January 22, 2024 | Posted in

A GWAS meta-analysis uncovers rare genetic variants, with large effects, associated with migraine. The research also pinpoints variants unique to different migraine subtypes.

Since publication of the first-ever genome-wide association study (GWAS) in 2005, researchers have performed thousands of such investigations, for thousands of unique diseases and traits, including for migraine. Those studies look for single nucleotide polymorphisms, or SNPs – genetic variations, at a single base in a DNA sequence, that are associated with a particular disease or trait of interest.

The majority of SNPs that GWAS uncover are common variants that confer only a small increase or decrease in disease risk. But rare variants may have much larger effects, and can give important clues about the pathophysiology of disease. A new GWAS meta-analysis now reports both common and rare variants associated with migraine, and in doing so offers new insight into different migraine subtypes.

Researchers led by Kari Stefansson, deCODE genetics, Reykjavik, Iceland, combined GWAS data from six European populations to identify four new SNPs that were associated specifically with the migraine with aura subtype, one of which was a rare variant that increased the risk of migraine with aura as well as epilepsy. The group also identified 13 SNPs associated specifically with migraine without aura, all but one in or near genes previously implicated in migraine.

Other rare variants with large effects also emerged from the study, several of which in one gene protected against migraine, and another variant in another gene that protected against migraine and brain aneurysms.

Together, the results provide new insight into the genetic basis of migraine subtypes and into their underlying disease mechanisms, and even suggest potential treatment targets.

“This is a very good study that builds and extends on previous GWAS,” said Lyn Griffiths, a molecular geneticist at Queensland University of Technology in Australia.

Griffiths, who was not involved with the new work, said that the strengths of the study include its focus on rare variants and on the migraine with or without aura subtypes. The use of large datasets, as well as proxy measures for migraine with aura and for severe migraine, also stand out, according to Griffiths (see MSC related interview).

The study appeared in the November 2023 issue of Nature Genetics.

Combining the data to learn more
Stefansson, along with first author Gyda Bjornsdottir, also of deCODE genetics, have a long-standing interest in migraine.

“Migraine is a very interesting disease. It’s not a plain old pain but rather a combination of so many things, and therefore migraine comes into the differential diagnosis of so many other diseases. It is fascinating to be able to find variants in the genome, but to ground diagnosis in a biological process is a particularly interesting thing to do,” said Stefansson, who founded deCODE genetics 27 years ago. The company has been studying migraine for about that long, too.

“We’ve been studying migraine for a long time and, like so many others, in terms of genetics, we have studied it as one disease,” said Bjornsdottir. “So, the approach now was to study, on a large scale, the main endophenotypes or subclasses of migraine,” she said, referring to migraine with and migraine without aura.

To do so, the researchers looked to six large GWAS that had reported SNPs associated with clinically defined migraine, migraine with aura, and migraine without aura. The data came from GWAS conducted at deCODE genetics on data from Iceland, Denmark, the UK Biobank, the US, and Norway, and the study also used summary statistics from GWAS conducted in Finland by the FinnGen study.

The researchers would perform meta-analyses in which they combined the data from all of these GWAS, giving them the ability to detect genetic associations that are otherwise hard to unearth in smaller, individual studies.

The group also performed meta-analyses of GWAS that had data about proxy phenotypes: visual disturbances preceding headache served as a proxy for migraine with aura, and bad and recurrent headaches were a proxy for severe migraine. That data came from individuals who answered questionnaires about whether they had such symptoms.

Together, the meta-analyses would include data from 1.3 million people, including 16,603 who had migraine with aura, 11,718 who had migraine without aura, 79,495 with any migraine, 30,297 with visual disturbances preceding headache, and 51,803 individuals with bad and recurrent headaches.

Distinguishing migraine subtypes
The meta-analyses revealed four new genetic variants associated specifically with migraine with aura. Of the four, a rare variant in a gene called PRRT2 had the strongest association with that migraine subtype; genes containing the PRRT2 variant have disrupted function. Notably, the PRRT2 variant was also associated with epilepsy, and had previously been associated with rare neurological disorders, including movement disorders, in case studies.

PRRT2‘s association with both migraine with aura and epilepsy touches on a long-standing discussion in the field about the nature of migraine, Stefansson said.

“The common denominator of the epilepsy and the migraine, in this instance migraine with aura, is the fact that both of them appear to be episodic, electrical dysrhythmias of the brain. And this brings us to the age-old debate about migraine: Is migraine just the spreading depression or electrical phenomena in the cerebral cortex, or is it primarily a vascular phenomenon?”

The researchers also found five variants associated with their migraine with aura proxy, four of which had not been reported before.

“Using these proxy measures is pretty novel, and it has allowed the authors to capture probably a whole bunch of people who wouldn’t have been diagnosed otherwise. So, bringing in the visual disturbances preceding headaches as a proxy for migraine with aura, for example, was a really good idea,” Griffiths told MSC.

Stefansson said that the results “unequivocally show that the biological basis of migraine with aura and migraine without aura do not completely overlap. There may be some overlap, but there is clearly a distinct biochemical difference between the pathogenesis of these two phenomena.”

With regard to migraine without aura, the group identified thirteen variants associated specifically with that subtype, most of which were in known migraine genes. The team also reported some interesting findings with regard to overall migraine. For instance, one variant was in the gene for nerve growth factor, a protein well known for its role in pain signaling.

Finding additional rare variants
The team found other rare variants, including in a gene called SCN11A, which makes an ion channel also very familiar to pain investigators for its role in pain signaling. The variants turned out to be associated with strong protection against overall migraine.

Another rare variant unearthed in the study was in a gene called KCNK5, which makes an ion channel that controls the flow of potassium in and out of the cell. The team uncovered this variant in one of their proxies and, interestingly, it also provided protection against brain aneurysms and coronary artery disease.

“This particular variant is quite an outlier in terms of the size of its effects, both on brain aneurysms and on bad and recurrent headaches, the proxy for severe migraine,” Bjornsdottir said.

As was the case with the PRRT2 rare variant, the KCNK5 rare variant also harkens back to the debate about the nature of migraine, here because of its association with brain aneurysms, by providing support for the vascular hypothesis of migraine pathogenesis.

“It’s not unreasonable to think that the KCNK5 variant is having an effect on both phenotypes, on the aneurysm and on the migraine, by affecting blood vessels in some specific way,” Stefansson said.

In light of the association of KCNK5 with brain aneurysms, the researchers speculated in their paper that perhaps that the beginnings of a brain aneurysm might sometimes be erroneously attributed to migraine instead.

Pathways and drug targets
Interestingly, based on their evidence implicating 22 genes as functionally important in migraine or its subtypes, the researchers used an additional approach, called pathway analysis, to identify biological pathways that may loom large in these conditions. A baker’s dozen of pathways related to nerve growth factor processing were among the most relevant pathways pinpointed in that analysis.

The new study also included results from a genetic drug target analysis. The team found that many drugs in development, and some currently used drugs, target the 22 genes. For instance, a drug called bryostatin, originally developed to prevent tumor growth but that is now in clinical trials to treat Alzheimer’s disease, targets PRRT2, the rare variant the investigators found associated with migraine with aura.

By pointing to drug targets that are the focus of current therapeutic efforts to treat other conditions, the study raises the hope that those drugs could perhaps be repurposed to treat migraine. That would save the time and money that developing new drugs from scratch entails.

What does the future look like?
While Griffiths praised the authors’ novel focus on rare variants, she said future studies could look at variants that are even rarer, such as private variants, which occur only in a single family or population.

She also noted the need to expand genetic studies to non-European or other unique populations.

“We need to be looking at more diverse populations, as there may be different variants that play a role in different populations,” Griffiths said.

She also recommended that future research extend these types of analyses to founder populations. A founder population is a small group of people who have become separated from a larger population. Because these smaller groups have reduced genomic variability, it becomes easier to identify genes that play an important role in disease.

Griffiths also said that artificial intelligence (AI) potentially has a big role to play in advancing future genetic studies of migraine, especially in terms of finding patterns in the data.

“One of the newer things that we should be thinking about is using machine learning and AI analysis, particularly for really large datasets like this, to see whether we can pull out more complex signatures rather than just a list of single genes – which genes are actually working together to create a profile.”

Showing the feasibility of such an approach, Griffiths pointed to one of her recent studies that combined machine learning with genomic data to explore outcomes related to head trauma. The approach made it possible to genomically distinguish those who had poor outcomes from those who had more normal recovery.

Other areas for future work that Griffiths recommended include examining epigenetic changes – alterations to DNA that don’t affect the underlying genetic sequence – which can be regulated by SNPs, as well as personalized medicine where each individual’s unique genetic profile can point the way to the best treatments.

As for the authors, up next is a sex-based analysis in collaboration with the International Headache Genetics Consortium that will look at the findings in women and men separately. “That will be a very important contribution because we know of the dramatic difference in the prevalence of the disease in the two sexes,” Stefansson said.

In the end, the study illustrates that when it comes to finding rare variants, bigger is better, according to Stefansson.

“The sequencing of large numbers of genes that is happening in genetic studies is not a futile effort. You need large sample sizes to find rare variants and also to have the power to demonstrate that they affect the risk of disease. Human genetics essentially is the study of human diversity – and it begins with the diversity in the genetic sequence.”

Neil Andrews is a science journalist and executive editor of the Migraine Science Collaborative. Follow him on Twitter @NeilAndrews

Image credit: 123RF Stock Photo.

Rare variants with large effects provide functional insights into the pathology of migraine subtypes, with and without aura.
Bjornsdottir et al.
Nat Genet. 2023 Nov;55(11):1843-53.

Sign Up For An MSC Newsletter

Neil Andrews is a science journalist and editor based in New York City. He has over two decades of experience covering science and medicine for expert and non-expert audiences alike. He is also the executive editor of the Migraine Science Collaborative, where he manages the day to day operations of the site. Previously he was the executive editor of the Pain Research Forum.

When not thinking and writing about neuroscience, Neil spends much of his free time on his Peloton and exploring NYC. He is also on a quest to satisfy his coffee cravings by visiting every independent coffee shop in the city. Follow him on Twitter @NeilAndrews.



  • Content Types

  • Paper Categories

  • Paper Dates

Recent News

Recent Papers


No event found!