Migraine triggers: how to find yours (and what a diary can and cannot tell you)

You get a migraine on a Sunday, after a hard week. Or after a bad night, or a long day with no lunch, or a wedding with loud music and a glass of wine. You start wondering what you did wrong.

This article explains what migraine triggers are, which ones have the best evidence, and why the picture is messier than most lists suggest. It also shows how to use a simple diary to learn your own pattern. It cannot tell you what is behind your headaches. Only a doctor can diagnose migraine.

What a trigger is, and what it is not

In headache research, a trigger is an exposure that raises the chance of an attack. It does not cause migraine by itself. Migraine is a condition of the nervous system, and the NHS says its exact cause is unknown. A trigger is more like a nudge on a brain that is already sensitive.

Two points matter from the start. First, triggers are individual. The American Migraine Foundation says that everyone has different triggers. Second, avoiding them is often not possible. A 2018 review by Marmura in Current Pain and Headache Reports notes that many common triggers are not easily modifiable, so avoiding them may not be realistic. You cannot cancel the weather or never feel stress.

What people report, and how strong the evidence is

The biggest summary so far is a 2018 meta-analysis by Pellegrino and colleagues in Cephalalgia. It pooled 85 survey studies with 27,122 people. About 81% of people with migraine or tension-type headache named at least one trigger. The most common were stress (58%), sleep (41%), emotion (33%), weather (32%), visual triggers such as bright or flickering light (32%), hormones (29%), food and eating habits (27%), smells (22%) and alcohol (21%). The authors also warned of extreme variation between studies, and said that how people were asked changed the answers.

That last point is important. These are what people believe sets off their attacks. Belief and proof are not the same thing. Here is how the evidence looks, trigger by trigger.

• Stress and the 'let-down' after stress: stress is the most commonly named trigger (Pellegrino 2018). In a 2007 clinic study of 1,207 people with migraine (Kelman), 75.9% reported triggers and stress was the one reported most often. A 2014 diary study by Lipton and colleagues followed 17 people for up to three months. Stress level on its own was not clearly linked to attacks. A drop in stress from one day to the next was linked to more attacks over the next 6 to 18 hours (odds ratios 1.5 to 1.9). The study was small (17 people completed more than 30 days of diaries, with 110 attacks), so treat it as a lead and not a rule. Marmura's review also lists this 'let-down' among triggers verified in clinical studies.
• Sleep changes and skipped meals: the Marmura review says clinical studies have verified fasting as a trigger. The NHS lists skipping meals, stress and tiredness among common triggers. Sleep is the second most named trigger in Pellegrino 2018. A 2026 systematic review (Sforza and colleagues) describes migraine and sleep disturbance as influencing each other in both directions.
• Hormonal change: periods are a well-supported trigger in women. Marmura notes clinical studies verifying the premenstrual period. See our article on menstrual migraine.
• Weather: a 2015 study of one-year diaries from 66 people in Taiwan found that temperature change explained part of the ups and downs in headaches in winter, and more so in people who said they were sensitive to temperature, mostly for milder headaches. Marmura says low barometric pressure is most likely a trigger. The study authors describe earlier research on weather and headache as inconsistent.
• Bright or flickering light, and strong smells: commonly named, but see the next section, because these may sometimes be early symptoms and not causes.
• Alcohol: a 2019 survey of 2,197 people with migraine in the Netherlands found that 35.6% named alcohol as a trigger, most often red wine. But red wine led to an attack every time in only 8.8% of people. The authors say this low consistency suggests that other factors also play a part.
• Foods such as chocolate: a 2020 review in Nutrients found chocolate named as a trigger by a small share of people in most surveys, and none of three placebo-controlled provocation studies found a difference. It concluded there is insufficient evidence that chocolate is a trigger.
• Dehydration and caffeine: the American Migraine Foundation says dehydration is a trigger for about one third of people with migraine, and lists caffeine as a trigger for some, while noting that for others a cup of coffee may ease symptoms. The NHS lists too much caffeine among common triggers. This is the weakest tier of evidence here, because these are survey-style statements and not trial results. In the one randomised trial we found (Spigt 2012, 102 people with recurrent headaches, not migraine only), drinking more water improved quality of life scores but did not change the number of headache days. It was small and not blinded, so it does not settle the question. Drinking enough water is sensible, but do not expect it to prevent attacks.

When a 'trigger' may be an early warning

A migraine attack does not start with the pain. Many people have a warning phase, called the prodrome or premonitory phase, hours or up to two days before the headache, according to the NHS. It lists early symptoms such as tiredness, yawning, food cravings, thirst, mood changes, a stiff neck and passing urine more often.

In a 2003 electronic diary study (Giffin and colleagues, Neurology), 97 people with usable data correctly predicted headaches from 72% of diary entries that contained warning symptoms. The people were chosen because they already noticed warning symptoms before attacks, so this does not apply to everyone. The most common were feeling tired and weary, trouble concentrating and a stiff neck.

Here is why this matters for triggers. If a craving for sweets starts in the warning phase, and you eat chocolate, the chocolate gets the blame when the headache arrives. A 2015 study of 1,010 people at a headache centre (Schulte and colleagues, Journal of Headache and Pain) found a clear overlap. People who named light as a trigger often had light sensitivity in the warning phase. People who named foods or smells as triggers often had food cravings or smell sensitivity in the warning phase. The authors concluded that commonly reported triggers are probably not so much independent causes as misread results of the warning phase. This was a retrospective study at one headache centre, and Marmura's review makes the same point more cautiously, that early symptoms may mimic triggers. It does not mean no trigger is ever real.

So a diary should record not only what happened before an attack, but also when the first odd feeling began.

Why a headache diary is the best tool

The NHS says a migraine diary can help you work out what might trigger your attacks. The American Migraine Foundation says it helps you and your doctor see patterns over time and decide whether you have migraine. There is also a quieter reason. Memory is unreliable. In a 2021 study of 484 people with migraine at a Dutch clinic (van Casteren and colleagues), the monthly migraine days people estimated differed from their diary by an average of about 4.7 days. Researchers reviewing trigger studies also say that prospective diaries, written as things happen, give the most reliable information (Pavlovic 2014, Headache).

What to write down

• The date, and the time the attack started and ended.
• Any odd feelings in the hours before: yawning, tiredness, craving, a stiff neck, mood change, light or smell sensitivity.
• How bad it was, on a simple scale of 0 to 10, and what else came with it, such as nausea or sensitivity to light and sound.
• Whether you saw flashing lights or zigzag lines, or had tingling, before the pain.
• What you took for it, when, and whether it helped. Write the name of the medicine, not just 'tablet'.
• The day before and the day of: sleep hours, meals and gaps between them, caffeine, alcohol, water, stress level, travel, weather.
• For women and people who menstruate: the first day of your period. See our menstrual migraine article.

How long, and how to read it

The NICE headache guideline says that if a diary is used, it should be kept for a minimum of 8 weeks, recording how often, how long and how severe the headaches are, any other symptoms, the medicines taken, possible triggers and how headaches relate to your period. For attacks linked to periods, NICE says the diagnosis should use a diary covering at least 2 menstrual cycles. Research diaries in the studies above ran for about three months.

When you read it back, look for things that repeat before attacks, not one-off events. Check the warning hours as well. Ask whether the 'trigger' shows up before attacks only, or also on many days with no headache. If you ate chocolate before three attacks but also on thirty days without one, the link is weak.

Testing a suspected trigger sensibly

These steps are common sense and are not a medical protocol. Change one thing at a time, so you know what made the difference. Try it more than once, as one attack proves little. Triggers can also work together. Even red wine did not bring on an attack every time in the 2019 study.

Be careful about cutting out many foods and drinks at once. Skipping meals is itself a reported trigger (NHS, Marmura), so a long list of banned foods can backfire. Strict rules can also add worry, and worry and stress are triggers too. If you think a specific food is a trigger, talk with your doctor, and ask about seeing a dietitian before removing whole food groups.

What to bring to a doctor

Bring your diary, or screenshots of it. Bring a list of all medicines you take for headaches, including how often. Say how headaches affect work, studies, family and mood. Mention when headaches started or changed. The NHS says to see a GP if attacks are severe or getting worse, happen more than once a week, or are hard to control. Manas also has a migraine diary at /migraine/diary where you can keep this record in one place.