Migraine JournalCaffeine and Migraines: A Two-Week Test You Can Run
Caffeine can ease an attack, cause one, or cause one by its absence. A two-week protocol that tells you which of those is happening to you.
Caffeine has three relationships with migraine, and two of them run in opposite directions. It makes painkillers work better, which is why it is an ingredient in combination headache tablets. In some people, at some intakes, it brings one on. And in anyone who drinks it daily, being late with it produces a headache of its own, listed as a diagnosis in the international classification. No article can tell you which of those describes you. Two weeks of deliberate logging usually can.
Why the same substance shows up on both sides
It helps a painkiller work. The Cochrane review by Derry, Derry and Moore pooled 20 studies and 7,238 people: adding caffeine at 100 mg or more to a standard analgesic gets about 5% to 10% more people to good pain relief, and the effect held for dental and period pain too. That is why caffeine-containing tablets sit on the pharmacy shelf.
It can also bring an attack on the same day. Mostofsky and colleagues (American Journal of Medicine, 2019) had 98 people with episodic migraine keep morning and evening diaries: 4,467 days, 825 migraines, each person compared against themselves. The association depended on habit. Among people who normally drank at least one serving a day, the raised odds only showed up on days with three or more. Among those who normally drank less than one, one or two was enough.
Going without it causes a distinct headache. ICHD-3 lists caffeine-withdrawal headache as 8.3.1: a headache developing within 24 hours in someone taking more than 200 mg a day for over two weeks, where intake has been interrupted or delayed. It is relieved within an hour by caffeine, or clears in seven days without it. In a small randomised crossover trial, Alstadhaug and colleagues took ten habitual heavy consumers off caffeine abruptly and triggered severe attacks in seven.
Read those together and the confusion makes sense. Someone whose morning coffee reliably clears a dull ache reads it as caffeine treating a headache. It might be. It might also be a withdrawal headache relieved by its own cause, which is criterion 8.3.1 working as written.
The weekend pattern, explained mechanically
The clearest case is the Saturday headache. In a study of 151 migraine clinic patients, Couturier, Hering and Steiner (Cephalalgia, 1992) found that patients with both a high daily caffeine intake and a markedly delayed weekend wake-up had a 69% risk of weekend headache. With only one of the two, 4%. With neither, zero. Their conclusion: “Sleeping in is not on its own a significant cause.”
The mechanism is arithmetic. First coffee at 07:00 on weekdays and 10:30 on Saturday puts you three and a half hours past your usual dose, on top of the overnight gap. ICHD-3 puts “delayed” alongside “interrupted” for exactly this. You did not quit anything. You slept in.
Delayed caffeine is one candidate among several, and the honest position is that they stack rather than compete — the full list of weekend suspects is worth reading before you settle on this one.
The two-week test
Change the timing before you change the amount. Timing separates withdrawal from trigger, and it is the variable nobody controls for.
Week one: hold steady, change nothing. Same drinks, same number, same clock times, weekend included. If you normally have two coffees at 07:15 and 13:00, that is what happens on Sunday as well, alarm and all. Each day, write down the number of caffeinated drinks, the clock time of the first one, whether an attack happened and how bad, and what you took for it. Log the quiet days too; a diary with only attacks in it has nothing to compare.
Week two: pick one arm, not both.
If your attacks cluster on weekends or days off, run the withdrawal arm. Keep week one’s timing through Saturday and Sunday: first cup at the same hour, even if you go back to bed afterwards. Weekend headaches that vanish under fixed timing are a strong lead.
If your attacks are scattered and you suspect volume, run the trigger arm. Reduce gradually across the week, times still fixed. Abrupt full cessation is itself a recognised cause of headache, so a cold-turkey week produces one whatever your real relationship with caffeine is, and tells you nothing.
The one-page trigger test tracker is built for this shape of test: a box for the single variable you are changing, weekly rows, and a column for anything else that shifted mid-test. That last column is what saves you.
Reading the result without overreading it
Fourteen days is a lead, not a conviction. Most people log three or four attack days in a fortnight, and three or four moves on chance alone. Two clean weeks buy a direction worth testing for longer.
Two traps apply, both covered in how to track migraine triggers without fooling yourself. The first is the premonitory phase: craving something, coffee included, can be the attack already starting hours before the pain. The second is changing more than one thing. A test week containing a house move is not a test.
Where your medicine cabinet complicates the picture
If you take a combination painkiller, the caffeine in the tablet is intake too, and those days count somewhere else. Combination analgesics sit at the ten-day-a-month line rather than fifteen, so check the drug-class-specific limits for medication overuse headache before attributing anything to your coffee. Twelve days a month of a caffeine-containing tablet is a bigger variable than a flat white.
Anything involving your medication, including reducing it, is a conversation with your GP or neurologist rather than a fortnight’s experiment.
Running it
Paper works: the tracker above is one page. If you already log attacks in Migraine Journal, caffeine is already one of the preset triggers; tick it and put the clock time of your first drink in the notes field, because the time is what makes the weekend pattern visible, not the tick. The rest of the tracking guides run off the same daily habit.
What you take into the next appointment is not “I think caffeine affects me.” It is “for two weeks I held my intake at the same times every day, weekends included, and my Saturday headaches stopped.”
This is background reading rather than medical advice. Caffeine-withdrawal headache and medication-overuse headache are both diagnosed by a clinician with your full history in front of them. If your headaches have changed pattern, or come on suddenly and severely, see a doctor rather than a diary.
Quick answers
Can caffeine both relieve and cause migraines?
Yes, which is why the advice is so confusing. A Cochrane review of 20 studies found that adding 100 mg or more to a standard painkiller gets 5% to 10% more people to good pain relief, while a 2019 diary study of 98 people found raised odds of an attack on days with more servings than usual.
Is caffeine-withdrawal headache a real diagnosis?
Yes. ICHD-3 lists it as 8.3.1: a headache developing within 24 hours in someone consuming more than 200 mg a day for over two weeks, where intake has been interrupted or delayed. It is relieved within an hour by caffeine, or clears in seven days without it.
How do I tell withdrawal from caffeine as a trigger?
Hold the amount and the clock times completely steady for a week, logging the quiet days too, then change exactly one thing in week two: delay the first drink, or reduce the amount. Changing both at once makes the result unreadable.
Track it in under a minute
Logs an attack in seconds, records the barometric-pressure change automatically, and prints the summary a neurologist asks for.
This article is general information, not medical advice. Talk to your GP or neurologist about your own symptoms, medication, and treatment.
