Quick Answer: Migraine is usually better described by a threshold model than a trigger model. Individual exposures each lower your resistance somewhat, and an attack begins when the combined load crosses a threshold that itself moves day to day. This is why the same wine causes an attack in one week and nothing in another, and why trigger diaries so often produce contradictory results. Managing the stack, and the baseline that sets the threshold, is usually more effective than eliminating any single item.

You tested the wine. Nothing happened. You tested it again a month later and lost a day.

So you concluded wine is not your trigger, or that it is unpredictable, or that you are imagining it. There is a third explanation that fits the evidence better.

What is the threshold model?

The trigger model says exposure causes attack. Simple, testable, and it does not match most people’s experience.

The threshold model says the migraine brain has a variable level of susceptibility, and that individual exposures add to a load. When the load exceeds the current threshold, the attack cascade begins. Two things vary independently: the load from exposures, and the threshold itself, which shifts with sleep, hormonal phase, stress recovery, illness, and factors nobody has fully mapped.

This explains several things the simple model cannot.

Why the same exposure produces different outcomes. The wine did not change. The threshold did.

Why attacks cluster. After an attack the threshold is often lower for a period, so a smaller stack sets off the next one.

Why triggers seem to appear and disappear over years. Baseline threshold shifts across life stages, particularly with hormonal changes, and a stack that used to be safely under the line stops being so.

Why elimination diets so often disappoint. Removing one item lowers the stack a little. If the rest of the stack is high, attacks continue and the elimination looks like a failure.

What typically goes into a stack?

The common contributors, roughly grouped, though the mix is individual.

Sleep. Both too little and too much, and irregularity matters more than duration. A weekend lie-in is one of the most commonly reported contributors and it is why weekend attacks are so frequent.

Hydration and meals. Skipped or delayed meals, and dehydration, both add load reliably and both are easy to fix.

Hormonal phase. For many people the days around menstruation involve a substantially lowered threshold, which is why a stack that is fine most of the month is not fine then.

Stress transitions. Attacks often arrive after the stressful period rather than during it, when cortisol falls. The high-stress week loads and the recovery day triggers.

Weather and pressure changes. Barometric shifts are strongly reported and largely outside your control, which makes them a good argument for managing the parts you can control on those days.

Sensory load. Bright or flickering light, extended screen time, loud environments, strong smells.

Specific foods and drinks. Alcohol, particularly red wine, aged cheeses, cured meats, and for some people caffeine changes in either direction.

Physical factors. Neck and shoulder tension, an unusually intense session, or the first cold morning after a warm spell.

Which matter most?

Sleep and meal regularity are the two most people underestimate, and they are also the two that raise the threshold rather than just reducing the load.

Myth vs Fact: migraine triggers

Myth: Everyone with migraine has identifiable food triggers. Fact: A substantial proportion never identify a consistent food trigger, and studies find food is a less frequent contributor than sleep, stress and hormonal factors. The focus on food is disproportionate to its role.

Myth: If something did not cause an attack this time, it is safe. Fact: Under a threshold model, safety is contextual. The same exposure is safe on a low-load day and not on a high-load day.

Myth: Chocolate is a trigger. Fact: Chocolate craving is a well-documented feature of the prodrome, the phase that begins hours before pain. Many people eat chocolate because the attack has already started, then blame the chocolate. This confusion appears in the research literature repeatedly.

Myth: You should eliminate every suspected trigger. Fact: Aggressive elimination produces restricted diets, social cost, and frequently no improvement, because it targets small contributors while ignoring the threshold. Most specialists advise against broad elimination.

Myth: A trigger diary will find the answer. Fact: Simple diaries frequently mislead, precisely because single exposures do not map onto single attacks. A diary that also tracks sleep, meals, cycle phase and stress is far more useful than one that tracks only foods.

How do I track this in a way that actually works?

Track the load, not the suspects.

• Log a small number of things daily, not just on attack days: sleep hours and timing, whether meals were on time, hydration, stress rating out of five, cycle day, and anything unusual.
• Rate each day’s total load, roughly, rather than looking for one culprit. Over eight to twelve weeks patterns emerge that single-item tracking will never show.
• Look at the 24 to 48 hours before an attack, not the two hours before. Prodrome symptoms mean the attack often started well before you noticed, and the causal window is earlier than instinct suggests.
• Watch for prodrome signals specifically: yawning, neck stiffness, mood change, food cravings, increased urination, difficulty concentrating. Learning yours gives you a warning window that is genuinely actionable.
• Do not test single triggers in isolation. A single-exposure test on a low-load day proves nothing, and on a high-load day it is not a controlled test either.

How long before this tells you something?

Usually two to three months. Attacks are infrequent enough that shorter periods produce noise, and the whole point of this approach is to see combinations rather than incidents.

What actually raises the threshold?

This is where the model becomes useful rather than just explanatory, because threshold is more modifiable than most people expect.

Sleep regularity above sleep duration. Same bedtime and same wake time, including weekends, is one of the most consistently effective non-pharmacological interventions. The weekend lie-in is worth giving up.

Regular meals. Not skipping, not going long gaps. This is unglamorous and it does more than most food eliminations.

Consistent hydration, spread through the day rather than concentrated.

Regular aerobic exercise. Reasonable evidence supports moderate aerobic activity reducing migraine frequency. Build gradually, since a sudden hard session can itself be a contributor.

Stress management with attention to the recovery period, since the post-stress drop is when attacks frequently land. Deliberately protecting the day after an intense period is a real strategy.

Consistent caffeine. Either amount is workable. Fluctuation is the problem.

Treating the underlying medical picture. Preventive medication, magnesium, riboflavin and CoQ10 all have varying levels of evidence and all are decisions to make with a clinician rather than from an article.

And the one people miss: reducing medication overuse. Frequent acute painkiller use can lower the threshold over time and produce medication overuse headache, which looks like worsening migraine and responds to a completely different intervention.

This is general information, not medical advice. Frequent or changing headaches should be assessed by a doctor, and any new, sudden or severe headache needs urgent medical attention.

FAQ: Trigger Stacking Questions, Answered

Does this mean I should stop avoiding my known triggers?

No. It means weighting effort correctly. Keep avoiding the ones with a clear, repeated pattern, and stop spending energy on marginal ones while ignoring sleep and meal timing.

Why do attacks come on holiday when everything is relaxed?

Classic post-stress pattern, usually combined with schedule disruption, travel, altered sleep and altered meals. Several stack items change at once.

If the threshold varies, can I predict bad days?

Partly. Cycle phase, poor sleep the night before, and known weather changes give a reasonable forecast. On predicted low-threshold days, deliberately reduce the controllable load.

Is one large trigger ever enough on its own?

Yes, for some people some exposures reliably cross the threshold alone. Those are worth avoiding outright. They are less common than the culture around migraine suggests.

TL;DR:

• Migraine fits a threshold model better than a simple trigger model: exposures add up, and the threshold itself moves.
• This explains why the same exposure is fine one week and not the next, and why single-item trigger diaries so often contradict themselves.
• Track sleep, meals, hydration, stress and cycle phase daily for two to three months, and look at the 24 to 48 hours before an attack.
• Raising the threshold through sleep regularity, meal regularity and aerobic exercise usually beats eliminating another food.

The most useful shift is from asking what caused this attack to asking what the load looked like that day. The first question rarely has an answer. The second one usually does.

This is general information, not medical advice.