SNPWatch & 23andMe Research Findings: Migraines and the Speed of Science

Editor’s note: This post has been changed from its original to reflect regulatory changes on what 23andMe can report to customers. Science is slow. It can take years to enroll patients in studies, run experiments and analyze data. To make matters worse, it can often take a year or more just to get those findings published. Wouldn’t it be great if there was a faster way? A recent paper on genetic associations in migraines gave our research team an opportunity to demonstrate the power of our unique research framework. In a study published last week in Nature Genetics, a large group of researchers led by Daniel Chasman of Brigham and Women’s Hospital in Boston identified three new SNPs associated with migraines. The researchers compared nearly 9,000 people who reported migraines to about 32,000 individuals who did not report migraines, all of European ancestry. They found   a single nucleotide polymorphism (SNP),   associated with 1.11 times higher odds of migraines, another SNP, was associated with 0.85 times the odds of migraines and another associated with 0.90 times the odds of migraines. Within a matter of days our research team ran an analysis of our own migraine cohort, comprising over 4,600 migraine sufferers and over 22,000 controls, and found similar results for all three of the associations reported in the paper. Of interest was the fact that the new Nature Genetics study found no association between migraines and , a link reported by a study published last year. And like Chasman’s group, we also found no association between and migraines in the 23andMe database. One possible explanation for these results may have been differences in determining if patients were suffering from migraines. All of the new patient groups in Chasman’s study self-reported whether they suffered from migraines, while the group used in the previous study consisted of clinician-diagnosed migraine sufferers. Here at 23andMe we also use self-report data, although our survey question asks specifically if the migraine was diagnosed by a physician. Further research is necessary to fully understand the role of genetics in migraines, however these new findings — replicated by our scientists — provide promising insight into the nature of the condition. Thanks to our customers science doesn’t have to be slow. Those interested in migraines can continue to help move the field forward by answering the migraine survey. There are also many other surveys covering a variety of topics that help our research team make exciting new discoveries. This is just one example of the power and speed of our research.
  • Love this, especially the takeaway: “science doesn’t have to be slow.” Citizen science can make swift strides in research when existing data sets are made available, and when the science-interested community is aware that they can “tinker” with the data.

    People who are motivated by a desire to make data free, to improve the medical system or by something entirely personal, like chronic illness, are natural citizen scientists who would like to make a difference through research. (Traditional) Science may be slow, but we have communities anxious to be involve. For the benefit of the world, it behooves us to engage them.

    There is an issue with citizen scientists not knowing how to design or run studies, or how to even contribute data, but that is simply a matter of education and engagement. Those who bridge the gap from traditional science to community can teach these skills, thus empowering a much greater amount of people and research, done almost “in the background.” It should not be meant to replace traditional research, but will serve to complement it.

    I appreciate the work of 23andMe, both its contribution to important scientific research and education/engagement with the public. Thank you!

  • manal


  • Tracy

    I am 50 yrs old and I have had chronic Debilitating migraine since the age of 17. No one on either side of my family has migraines . This study is of great interest to me but I need more information.

  • Susan Price Davis

    When I did the survey, it told me that I was unlikely to suffer from Migraines. But, I have migraines, have since the age of 11 (I am 47 now). I have migraine with aura, have had hemiplegic migraines and ocular migraines (without the headache). I also get focal seizures. I am concerned that your survey did not capture me correctly and therefore, I believe you may have skewed information. I am however CC for snp rs2651899. My son who has had migraines is also CC. My father who reports only a few ocular only migraines is heterozygous. Anyway, I think your survey is not capturing things correctly.

    • Scott23H

      Thanks for the note. The survey is looking at one factor that influence the risk for migraines – lifestyle, health history, etc. The test looks at your genetics. Together they are supposed to help you assess your risk. A risk is simply that it doesn’t mean that you will or will not get a condition. As for the survey it was developed to identify and attempt to classify the types of headaches you get, their intensity and based on your responses (looking at triggers), the likelihood you’ll experience migraines. Obviously it’s not perfect. We’re always trying to improve the surveys so thanks for the feedback.

  • 23blog

    Hi KRMU,
    You are correct that originally there were specific SNPs referenced that we had to remove because of regulatory issues. This post has been changed to reflect that. If you can, I recommend going directly to the paper itself to find the specific SNPs.