Ask people with migraine what sets off their attacks and you get a long list: chocolate, cheese, weather, wine, stress, skipped meals, bright light. Some of these hold up when researchers test them properly. Others fall apart, and one of the most famous triggers may have the cause and effect backwards. This post sorts popular triggers by how well they survive controlled study, so you can spend your effort avoiding the ones that matter and stop policing the ones that do not.
Why self-reported triggers are unreliable
Most of what people "know" about their triggers comes from looking back after an attack and picking a likely culprit. That method is wide open to bias. You remember the chocolate you ate before a migraine and forget the times you ate chocolate with no migraine. Trigger lists handed to patients also seed the answers, so people report the triggers they were told to expect.
The stronger evidence comes from two designs: prospective diary studies, where people log exposures and headaches forward in time, and provocation studies, where researchers deliberately give the suspected trigger or a placebo under blinded conditions. When a "trigger" survives those, it is probably real. When it only shows up in after-the-fact recall, it is suspect.
Triggers that hold up
Several triggers have consistent support across better-designed studies.
Stress, and the let-down after stress. Stress is among the most reliably reported triggers, and it holds up in prospective work. The pattern includes the post-stress let-down, attacks that hit as tension eases, like the start of a weekend after a hard week.
Sleep disruption. Both too little and disrupted sleep, and sometimes oversleeping, track with attacks in diary studies. Altered sleep patterns rank consistently high as a non-dietary trigger.
Skipped meals and fasting. Going without food shows up repeatedly as a trigger, plausibly through blood sugar and the stress response. This one is actionable: eating regularly is a low-cost change with real support behind it.
Menstruation. This is one of the best-established triggers, with a clear mechanism. Attacks cluster just before and during menstruation, driven by the drop in estrogen in the days beforehand. The hormonal timing is consistent enough that menstrual migraine is a recognized pattern with its own management.
Weather changes. Weather, especially shifts in barometric pressure and storms, holds up reasonably well in prospective data, though the effect size is modest and varies between people. Rain and storm conditions rank among the higher non-dietary triggers in tracking studies.
These are the triggers worth taking seriously, because the evidence behind them is more than recall.
Triggers that fall apart, and the chocolate trap
The most instructive case is chocolate, the trigger almost everyone names. When tested under blinded provocation, it does not hold up. A double-blind study gave participants either chocolate or a carob placebo and found chocolate was no more likely to provoke a headache than the placebo (Marcus et al., Cephalalgia 1997). Across provocation studies, there is insufficient evidence that chocolate triggers migraine.
The reason is a reverse-causation trap. Imaging shows the hypothalamus activates in the early premonitory phase of a migraine, before the headache (one of the features that separates migraine from an ordinary headache), and that phase can drive food cravings, including for sweets. So people crave and eat chocolate because a migraine is already beginning, then blame the chocolate when the headache arrives. The chocolate is a symptom of the oncoming attack, not its cause. This is the single most important idea in trigger science, because the same logic likely applies to other "food triggers" people swear by.
This does not mean no food ever matters. Some foods have plausible mechanisms: aged cheeses and red wine contain tyramine, histamine and sulfites, and red wine in particular carries several of these at once. Aspartame is labeled a possible trigger. But the evidence for specific dietary triggers is far weaker and more individual than the confident lists suggest, and chocolate is the cautionary example of how a craving gets mistaken for a cause.
How to find your actual triggers
Given how unreliable recall is, the honest method is to test prospectively rather than assume.
Keep a headache diary for a couple of months. Log attacks and the candidate exposures, sleep, meals, stress, menstrual timing, weather, specific foods, going forward, not from memory. Patterns that show up consistently in the log are worth acting on. A "trigger" that you assumed was real but never lines up in the diary probably is not yours.
Be especially skeptical of food triggers that appear in the hours before an attack, since the premonitory craving phase makes those the most likely to be reverse causation. And resist the urge to eliminate many foods at once, which shrinks your diet for little benefit and can add its own stress, itself a better-supported trigger than most of the foods being cut.
The bottom line
Stress and its let-down, disrupted sleep, skipped meals, menstruation and weather changes hold up across better-designed studies and are worth managing. Chocolate, the most famous trigger, largely fails blinded testing and is probably a victim of reverse causation, craved because the migraine is already starting. The reliable way to find your real triggers is a prospective diary, not after-the-fact blame. Spend your effort on the triggers the evidence supports and stop fighting the ones it does not.
Medically reviewed by: [Reviewer name, credentials] — [Date]
If trigger avoidance is not controlling your migraines, a licensed clinician can discuss acute and preventive treatment options. Start an online visit.
This article is for general education and is not a substitute for personalized medical advice from a licensed clinician.