My n=1 Quest to Live Headache-Free

I started getting headaches as a teenager, and migraines when my first daughter was born 9 years ago. I’m 34 now, so that’s about 20 years of some kind of regular head pain, and who knows how many ibuprofen pills popped.

When I started seriously self-quantifying a couple of years ago, pain was high on my list of things to measure and hopefully eliminate through n=1 experimentation. It’s been a long, slow process of tiny, incremental insights.

Here’s what I did, how I did it, and what I learned.

July 2009 – I started tracking my detailed food intake with the DailyBurn iPhone app. I did this for almost a full year, and started to get a good sense of how I felt after eating different kinds of food.

I thought I noticed a pattern that changes in my dairy intake caused migraines, especially having a lot of cheese or butter one day and then none the next day. I consciously tested this hypothesis after reading a post on Seth Roberts’ blog about butter – I ate 1/4 stick of butter, and sure enough, got a splitting migraine.

CureTogether’s list of migraine treatments gave me the idea of wheat/gluten being a possible offender, so I tested this too, and it was.

July 2010 – By now I had enough evidence that both dairy and gluten were triggers, so I gave them up entirely. I had 3 weeks of intense withdrawal headaches, and then an almost miraculous experience of no pain. I started using the Countdown iPhone app to record headaches and experiments.

Also, I noticed a very distinctive signature for dairy vs. gluten headaches, as graphed above. If I was accidentally exposed to dairy (sometimes the Starbucks baristas forget that you asked for soy), I would get a strong headache the next time 1 pm came around, which would last 6-8 hours. If I was accidentally exposed to gluten (like before I learned that soy sauce has wheat in it), I would get a strong headache the second time 5 pm came around, which would last 4-6 hours. This is highly repeatable in me.

August 2011 - I started noticing more frequent headaches again, but now I had a year of headache data, so I could actually chart it!

Some notes on this chart: GFDF is when I went gluten-free and dairy-free. There was still a background level of 1-2 headaches a month, one of which can be attributed to my menstrual cycle, the other to accidental dietary exposure. In October and February I actually had no headaches strong enough to record (if I have to take medication for the pain, I record it in Countdown).

The recent increase in headaches was puzzling, especially since they were getting stronger again. A couple of weeks ago, when I was having a clear migraine (my first one in over a year), my husband started scouring the web for ideas. He came across Seth Roberts’ post on cleaning products triggering migraines. Then he remembered that just the other day, he had bought some Febreze and sprayed our whole house with it. Could cleaning products be triggering headaches?

I went back through my Google Calendar, and sure enough, every time the cleaning lady came, 2-3 days later I would record an unexplained headache in Countdown. So we asked her to switch to only vinegar, baking soda, and unscented detergents that we provide for her. Last time she came, I didn’t get a headache!

Other offenders I’ve come across are:

Caffeine and sugar – I recently gave up chocolate and my decaf soy lattes, and all other sugar for good measure. I noticed that every time I went to a conference, I would eat a lot of chocolate and end up with a strong headache when I stopped. Some might say, why stop, just keep eating it! Well, then I have to be super careful to regulate the amounts to make them constant, and it doesn’t make my body feel good in other ways. So it’s just easier to eliminate it entirely. And when pain is a motivator, it’s not hard to stay on track. I feel much calmer and more balanced off sugar, too.

Gum – During stressful periods, I used to go through 2 packs of gum a day. The physical act of chewing would make my jaw hurt, which would turn into a headache. Gum is now on my blacklist, too.

Shoulder pressure – Heavy backpacks are a trigger, and when I get a massage, I always make sure to tell the person to not put too much pressure on the muscle between my shoulder joint and my neck (the superior trapezius muscle, if you want to get jargony).

Menstrual cycle – I usually get one headache just before my period, which must be hormonally related. No idea what to do about this one yet.

Emotional intensity - If I have a day where I cry a lot, say if someone dies or there’s some kind of major loss, it can trigger a headache. Being off sugar helps me to stay more emotionally balanced, but this does happen occasionally.

So that’s everything I’ve learned about my headaches so far. I’m investigating a couple of other suspects, like tyramine and blood sugar, but I don’t have anything to report on them yet.

What do I want from all this? It would be awesome to go a whole year without a headache – but that’s a soft intention to move towards, not a cold, hard goal. I hope that sharing this story of self-experimentation helps some of you out there with your own processes. Thanks for reading!

This entry was posted in Personal Projects and tagged , , , . Bookmark the permalink.

19 Responses to My n=1 Quest to Live Headache-Free

  1. Annie says:

    That’s so amazing! I’m glad you found something that works for you. Did you used to have a constant headache, or did you get them intermittently? I’ve had a constant headache for the past eight years, and I’ve tried different diets several times. How long did it take you to cleans your body of all the triggers?

  2. John Sumser says:

    That’s really helpful, Alexandra. It’s wonderful to get insight into your approach and results.

    As I dig into my own data, I’m becoming persuaded that everything I know about having, using and measuring my body is pretty primitive. It’s like a black box where I can (as you’ve nicely demonstrated) get a strong grasp of some of the input and output variables. The middle part, where my internal processes operate out of sight and away from measurement, is what I’m really trying to understand.

    I’ve been reading a great book, When The Body Says No by Gabor Mate. He does a good job of describing our bodies as tightly integrated neuro-immuno-hormonal-emotional systems. The idea that these things are entirely separate spheres gets in the way of my ability to understand the workings of my mind and body. I think you’d find it interesting.

  3. Alexandra Carmichael says:

    Annie – Thanks! I’ve always had them intermittently. Having a constant one sounds like a bigger challenge! It took a good 3 weeks to get the wheat and dairy out of my system.

    John – I loved your talk at the last QS! It made me more aware of really focusing on reducing stress where I can. Thanks for the book recommendation – it does sound interesting, and I’ve just wishlisted it!

  4. Wow – I can totally relate! Myself, I’ve often noticed that I get headaches after using my cellphone for an hour (as it turns out, my cellphone emits more radiation than most).

    And also – I get headaches when I’m sitting in the back of the bus for too long, and when I’m walking in a big city for too long. Apparently, air pollution can be really nasty on the brain – see

    Also – I get headaches after eating too many high-glycemic foods.

    I wonder if anyone else can relate to that

  5. pc says:

    The NYT just had an article about how dehydration can be a contributing factor in migraines. Good to keep in mind during these hot summer months!

  6. Because this trial study used only one subject, coincidence cannot be ruled out.

    > I ate 1/4 stick of butter, and sure enough, got a splitting migraine.

    We cannot soundly conclude that because eating butter was followed by a migraine, that the butter must have caused the migraine. Just because one event follows another event, that does not necessarily mean the first event caused the second event. There are other possibilities: it could be coincidence, or a third, underlying factor may have caused both events.

    Ruling out coincidence requires using multiple subjects.

    And the author’s trial was not blinded so did not rule out the placebo effect. (Or reverse placebo effect, in this case.) Placebo has a remarkably profound effect on headache.

    The author mentions dairy, gluten, Fabreze, caffeine, chocolate, sugar, gum, shoulder pressure, menstrual cycle, emotional intensity, tyramine, and [low] blood sugar as possible causes or triggers of her migraines.

    We can sort those factors into the following categories:

    Never been demonstrated to cause headache: dairy, gluten, Fabreze, sugar, gum, tyramine, and shoulder pressure.

    Infused tyramine was associated with headache in at least one study, but ingested tyramine has never been demonstrated to cause a headache.

    Demonstrated to cause severe headache and nausea/vomiting: Caffeine.

    Chocolate contains a lot of theobromine, which is similar in molecular structure to caffeine. Chocolate also contains some theophylline, which is also similar to caffeine.

    Promotes migraine: female gender and female reproductive events including menses.

    Symptom of migraine rather than cause of migraine: strong emotions.

    Low blood sugar can cause headache, but a headache caused by hypoglycemia is not a migraine. Also: hypoglycemia is quite rare among non-diabetics and normal-sized adults. You can easily test whether a headache is caused by hypoglycemia by drinking sugar water or fruit juice: if the headache isn’t relieved within 15 minutes, it wasn’t caused by hypoglycemia.

    The popular belief that just about anything and everything can cause (or “trigger”) a headache, and that identifying and avoiding your personal headache triggers is an effective treatment strategy for headache and migraine, is largely mistaken. No food has ever been demonstrated to cause headache, and dietary limitation is not an effective treatment for headache or migraine.

    We should concentrate on those factors demonstrated to cause headache. Among the factors the author names as suspects, only caffeine has been demonstrated to cause headache. Caffeine is therefore the most likely culprit.

    Caffeine readily explains premenstrual headaches. Estrogen inhibits caffeine metabolism. In the days prior to menses, estrogen body concentration decreases, so the rate of caffeine metabolism therefore increases, and with it the risk of caffeine withdrawal headache.

  7. Alexandra Carmichael says:

    Wow, Barry, that’s an interesting comment, but just because something hasn’t been shown to prove headaches in traditional research doesn’t mean it doesn’t cause headaches *in me*. Especially the dairy/gluten headaches – those have been repeated dozens of times, accidentally.

    Also, Seth Roberts posted a great perspective on this over on his blog –

    • Annamarie Castrilli says:

      Hello Alexandra,

      let me first commend you on your research. I too have suffered from migraines all my life and can manage them only because just like you I am constantly learning about what works and what doesn’t for me.

      Also I wanted to comment on the message from the previous contributor to this space. I, like you, do not view traditional research as having all the answers – quite the contrary. There is no doubt that for me certain foods and substances trigger headaches. These include chocolate, red meat, excessive carbohydrates (probably because of the sugar and gluten content), sweets, MSG, and legumes. Dust and heavy fumes also have the same effect.

      I doubt that the previous writer would say that the headaches are not real because traditional research does not have empirical proof that they do.

      Congratulations on your work and that of CureTogether which, just as its name implies, is dedicated to finding solutions among the community of people that share similar symptoms. The work you do there is incredibly important and gives hope to all of us who know full well that traditional research has failed us.

      Thank you so much!

  8. Alexandra and Annamarie: first of all, nobody is saying the headaches aren’t real. The headaches are definitely real.

    Alexandra, you’ve noticed an association between eating gluten or dairy and getting a headache later. Annamarie, you’ve noticed an association between eating chocolate, red meat, carbohydrates, MSG, legumes, dust, and heavy fumes and getting a headache later.

    The first question we need to ask is: are the associations you’ve noticed real or apparent?

    People often make the error of selective observation: they notice positive associations, such as headaches after eating rutabagas, but fail to notice, remember, or record non-positive associations, such as headache but no rutabagas, rutabagas but no headache, and no headache nor rutabagas. Those non-positive associations may not seem as remarkable as a positive association, but are just as important for the purpose of distinguishing between real and merely apparent associations due to observational bias. So we have to take observational bias into account.

    If we are able to rule out observational bias and determine the associations are real associations, the next question we need to ask is: are the associations causal or coincidental?

    We can’t soundly conclude that because eating gluten was followed by a headache, the gluten must have caused the headache. Could be coincidence.

    Even if eating gluten is on multiple occasions followed by headache, that still wouldn’t tell us that gluten must have caused the headaches. It could still be coincidence.

    Headaches are commonplace, and most people eat several times a day, so many people will notice associations between eating foods and getting a headache later. Even if no foods cause headache, random chance dictates that headaches will often follow the eating of this or that food.

    So: that’s where the clinical trial study comes in. The clinical trial study is a way to rule out random chance and demonstrate that A has a real effect on B.

    There’s a very popular and well-entrenched myth that foods trigger migraines, yet foods cannot be demonstrated to trigger migraines. And dietary limitation, it turned out, doesn’t help migraine patients beyond a placebo effect. So it’s well past time for everybody to discard the food trigger theory, which has been demonstrated to be wrong. I don’t want migraine patients wasting their time and energy trying to identify and avoid trigger foods, when it’s already been determined foods don’t trigger migraine.

    To identify the true causes of and cure for migraine, we must use good reasoning. If we have a hunch that, say, gluten triggers migraines, we must test our hunch. Clinical trial studies are a way to test our hunches.

  9. Alexandra, you gave up your decaf soy lattes. Do you drink coffee in other forms? Or tea of any kind?

    You note that when you ate a lot of chocolate then stopped, you got a strong headache. I think that tells us what causes your headaches: caffeine withdrawal. (Chocolate contains theophylline and theobromine, close chemical relatives of caffeine.)

    • Alexandra Carmichael says:

      Hi Barry, no, I don’t have any caffeine now. I wonder why you believe the part of my story about caffeine, and not the part about gluten – could there be some bias on your part too? Also, I’m not suggesting that my results apply to anyone but myself. The point of this whole post was to describe a method of self-experimentation, not focus on any particular result. I guess I’m feeling frustrated.

  10. Daniel Reda says:

    Actually, the placebo effect can be ruled out here. If someone has no knowledge of being exposed to a trigger prior to experiencing a headache, it can’t be placebo.

    In fact, this is even superior to a blinded placebo-controlled trial, where the subject knows they are ingesting something and believes it may have an effect. Exposing someone to a substance that may cause them harm, without their knowledge, would not even be allowed by an ethics committee, so blinded, placebo-controlled is the best you can do in a planned clinical study. You can learn a lot from situations where such complete blinding happened accidentally.

    If you believe gluten causes your headaches and eat sushi, expecting no headache because you believe it does not contain gluten, then experience a headache, then go back and discover that the #1 ingredient in soy sauce is wheat, it’s particularly strong evidence supporting your hypothesis that gluten triggers your headaches.

    Being surprised by an observation that conflicts with your expectations is powerful evidence because it rules out confirmation bias – it shows that you’re not just waiting for evidence to confirm your preexisting belief, which would more likely occur by chance alone.

    One way to establish causality is if you notice a dose-response relationship. In Alexandra’s case, her headaches were not merely coincidental with exposure to household cleaning agents, but were more frequent and more intense with greater exposure.

    If you suspect a confounding variable – e.g. the same thing that triggers your headaches also gives you a craving for dairy – then you can rule this out by scheduling ingestion (ideally randomly) of dairy and seeing if it causes a headache. Even better is if you also randomly vary the amount of exposure to see if there’s a dose-response relationship.

    Yes, subjective experience is often subject to a variety of biases, but this does not mean you can’t do good science as an individual.

  11. Seth Roberts says:

    Barry, I have never heard or read anything like “Because this trial study used only one subject, coincidence cannot be ruled out.” John Tukey, in this paper,

    pointed out that all studies are n = 1 in various ways: one classroom, for example. Perhaps you are confusing subjects with observations. Alexandra had many observations.

    Can a single observation rule out coincidence? Yes, it happens all the time. In practice, “ruled out” means “becomes highly implausible”. Depending on the details of the observation (how close in space and time are the coincident events?), their base rates, and prior beliefs (how plausible are alternative explanations?), coincidence may indeed be ruled out. In the case of Alexandra’s butter observation, the ingestion of the butter and the beginning of the headache were close together in time, the base rate of headaches was relatively low, and the plausibility of an alternative explanation (dairy causes headaches) relatively high. It was perfectly reasonable to rule out coincidence based on a single observation. Search Reichenbach’s Principle to learn more about how coincidence is ruled out from single observations.

  12. Eric Moyer says:

    You may be able to get rid of many of the menstrual headaches by talking with your gynecologist. It is possible to use hormonal birth control to eliminate or reduce the frequency of menstruation. I know (thanks to the magic of Google) that in 2007, “The Food and Drug Administration approved Lybrel, the first birth-control pill explicitly designed to abolish monthly bleeding.”

    I wish you well on your quest.

  13. Lisa Morrell says:

    Another thought is that headaches can sometimes be the result of a series of reactions in the body, not only a direct result.

    My husband is a cancer patient and experiences all sorts of headaches. At one point he had a migraine that persisted for several days, until we found out that he had a staph infection and started treating it – at which point the migraine went away. Our theory is that headaches are his body’s way of letting him know that something is off (or that he’s had MSG or Red Dye 40…). There have been times that headaches are not listed as a symptom or side effect of a condition or drug, but my husband gets them anyway while taking the drug or concurrently with the condition.

    So certain foods may be triggering an allergic reaction – or some other problem – in Alexandra that triggers a headache. There are many people who experience rare reactions to foods and drugs – the more lethal ones are noted, but headaches often go unstudied.

  14. Noumenon says:

    The many different things you’ve had to stop doing remind me of the many different postures and activities I avoid due to musculoskeletal pain. It would be great if we could fix the underlying causes, but just being able to avoid the pain is enough.

  15. Pingback: Modeling Diary Studies After Smartphone Apps | SurveyPost

  16. Lance Roberts says:


    I quit caffeine completely 6 years ago. My overall migraine frequency has not changed (still one/two per year), however I do get less headaches now (caffeine does give me headaches almost immediately if I don’t consume the same amount every day). I used to get a lot of headaches, and caffeine withdrawal does give me a similar headache to migraine (e.g. can include nausea and vomiting), but I can confidently say that there is no relationship between caffiene and full aura migraine episodes for me.

    My migraines come on with extreme stress (e.g. falling into the cold sea, or overexerting in a sports game), and follow around 24 hours later. They begin with a 20 minute period where a blurry pattern (e.g. drop some water on your computer screen and see the colours change) grows increasingly in my peripheral until it ceases and bam the pain hits.

    I have not had my blood tested for unexpected sources of caffeine coinciding with these events however I am well aware that caffeine is in many things and actively avoid all of those things.

    I think caffeine should be regulated R18 and the “aspirin” industry will suffer as people wake up to the amount of headaches that caffeine causes.

    I applaud calls for more research into headaches and their causes, however the pills treating it are where the money is as human nature seems to be that avoiding things is too much hassle.

    I think trying to claim that migraine doesn’t exist as distinct from caffeine withdrawal is not supported by medical research or evidence and is simply regarded as a side show by serious researchers. Certainly it doesn’t explain everything to me – I already fully believe caffeine is terrible and gives me many many headaches, and I avoid it at all costs and any lapses are always punished, however I can’t avoid the occasional actual migraine due to other causes.

    Hope this helps


  17. Val says:

    Nice post, thanks

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Notify me of followup comments via e-mail. You can also subscribe without commenting.