Jamie Scott sent me a link a couple days ago to this science daily article:
First Genetic Link to Common Migraine Exposed
So I chased down the study, as the article mentions some of my favorite words, glutamate and synapse. And it is a cool study! You've got to love the large, population based genetic studies they can do now, because they get p values (which is (sort of*) the chance that the finding occurred by chance) of 5.38X10-9. After scouring the nutrition literature for a while, it has been nice to read up on neurologists and geneticists who are gleefully obsessive rather than the nutrition epidemiologists who try to play Jedi mind tricks. "Bran is good for you. Also whole wheat. You will eat bran! You will!" I mean, I expect a pharmaceutical company to try to con me a little. They're in it to win it, after all. Where's the fun in reading another drug trial except to find the underdosed competitor medicine or the fouled up control? But it's no fun when the nutritionists do it. Especially with my tax dollars. I'd rather have a iRobot 330 Scooba Floor-Washing Robot, really, with the money.
But back to the study! I'll just gush a little, because you can certainly go read the Science Daily article yourselves. These geneticists took DNA from 3,279 headache-afflicted people from very nice countries with socialized medicine and possibly no need to buy disability insurance where they feel cool about giving up DNA to the government. They also took DNA from 10,747 matched controls. Then they used magical DNA replication and reading machines to roll out the genes for all these people, to find which ones the migraine sufferers had in common, and the controls didn't. Turns out it there was only one (of significance). The minor allele A of marker rs1835740 (p=5.38X10-9).
Genes have been found for migraines before, but always in rare family clusters with somewhat bizarre ion channel issues, and none of those genes were ever found in a large number of average migraine sufferers (8% of men and 17% of women). So to find a gene by splashing everyone's DNA out on a big canvas and finding the pattern is pretty big news! The researchers did all sorts of cute tests involving geneticists' favorite words, like HapMap and ssSNPs ("snips"), and no matter how they spliced the data, the same gene (rs1835740) came out as the common migraine one (not in every headache sufferer, to be sure, but in many!)
This is where geneticists get the full thumbs up Awesome. They went out and recruited another 3,202 cases (including some from some different socialized medicine countries), and 40,062 more matched controls, and they did the analysis again. And found that the same gene was over-represented in the migraine group, this time with p=1.69X10-11. Heh.
The punchline. rs1835740 is an area of a chromosome that has two genes for glutamate regulation. Yes, glutamate, that excitatory neurotransmitter that can be exceedingly annoying and cause all sorts of trouble (like seizures, bipolar disorder, depression, and migraines) when the regulation is out of whack. The actual gene they think is implicated is MTDH. MTDH is responsible for downregulating the major glutamate transporter in the brain.
The hypothesis of migraines is that too much glutamate is left out in the synapse, causing too much excitement in the wrong place at the wrong time, leading to spreading neurotoxic communication, head pain, sometimes aura - a migraine. Why would too much glutamate be left out in the synapse? Because some people appear to have inefficient pumping mechanisms to get it back into the cell. The glutamate transporter is one you need to be working tip top!
This is all indirect evidence, but it is sensible and very cool. Maybe your common migraines are due to this very gene and mechanism. Perhaps topamax or valproate or other GABA-influencing medicines could work to improve the headaches. Or you could actively work to reduce your stress so the glutamate isn't so prevalent. Or maybe even try a ketogenic diet. (not an FDA approved treatment for migraine - and I couldn't even find any case trials on pubmed, but I have heard of cases mentioned on the internet. I'll look harder) Intriguing!
* I've been called out - here's the precise definition of the p value -"In statistical hypothesis testing, the p-value is the probability of obtaining a test statistic at least as extreme as the one that was actually observed, assuming that the null hypothesis is true."