Perimenopause Migraines: Why They Get Worse First

The thing nobody warns you about is the order. Attacks often get worse for several years before they get better, and the worse part arrives while your periods are still happening — which means most people spend that stretch assuming the condition itself has deteriorated.

The mechanism points the other way. It is not low oestrogen that associates with attacks so much as changing oestrogen, and perimenopause is the period of maximum change. The Migraine Trust’s page on migraine and perimenopause sets out that oestrogen levels mirror changes in migraine prevalence, with fluctuating levels during perimenopause increasing the likelihood of both migraine and menstrual migraine.

Some numbers give it a shape. In the data the Migraine Trust cites, high-frequency migraine — more than ten days a month — was reported by around 8% of women still having regular periods, and by over 12% of those in perimenopause. In their survey of women attending a specialist menopause clinic, migraine affected 42%, a fifth of whom had daily headaches.

The measurement problem nobody mentions

Here is the practical trap, and it catches people who have been tracking diligently for years.

The standard method for telling whether attacks are hormonal is cycle-relative. Mark the first day of bleeding, mark the day each attack starts, and count how many attacks fall in the five-day window from day −2 to day +3, across three consecutive cycles. Two out of three is the pattern. How to tell if your migraines are hormonal works through the arithmetic, including the fact that there is no day 0.

That method quietly stops working in perimenopause, and it fails silently rather than obviously.

It needs three comparable cycles. Perimenopause gives you a 24-day cycle, then a 41-day one, then nothing for two months, then two in five weeks. There is no stable interval to anchor a window to, so a real hormonal pattern gets scattered across the count and reads as noise — and the conclusion “it’s not hormonal after all” arrives at precisely the point when hormones are doing the most.

What to record instead

Stop counting relative to the cycle. Start counting per month.

Keep marking two things. The first day of any bleeding, however irregular or unexpected, and the start date of every attack. Both marks stay useful later, and reconstructing either afterwards is impossible.

Switch the question to a rate. Migraine days per month, total headache days per month, plotted across twelve months rather than three cycles. A rate needs no cycle to anchor to, which is exactly why it survives the transition.

Add the cycle length. One number per bleed: days since the last one. Over a year this shows you the shape of your own transition, and it is the context that makes the migraine rate readable — a run of shortening cycles alongside a rising attack rate is a very different picture from a rate that rose while cycles stayed regular.

Three marks, none of them taking more than a few seconds. The free printable diary has a monthly grid with a box per day, which is the right shape for all three, and Migraine Journal keeps the migraine-days-per-month figure on a rolling window so the rate is a lookup rather than a monthly counting exercise.

Why the record matters more in this stretch than any other

Two reasons, and both are about decisions being made in the next few years rather than about curiosity.

The first is that hormone therapy, contraception and migraine interact in ways that depend on specifics — particularly whether you get aura, which is not a detail to establish from memory in an appointment. Whether any of it applies to you is a conversation with a GP or a menopause specialist, and it is a conversation that goes considerably better with dates in front of both of you. There is nothing on this page, or any page, that should substitute for that.

The second is that this is the stretch where attribution gets genuinely hard. Sleep is disrupted for reasons unrelated to migraine. Stress is often high for ordinary mid-life reasons. Hot flushes interrupt nights. Any of those can raise attack frequency on its own, and all of them arrive at once. Without dates, the whole period compresses into “everything got worse”, which cannot be acted on. With them, you can at least see whether the attack rate moved before or after the sleep did.

Aura is the detail worth being precise about

If you get visual disturbance, numbness, or difficulty finding words before or during attacks, record it explicitly rather than folding it into the severity.

Aura changes the clinical picture and it is one of the specifics that hormone-related decisions turn on. It is also something people under-report, partly because a silent aura with no headache afterwards does not feel like a migraine at all. Write down what you saw or felt, on which side, and how long it lasted.

What to expect afterwards

For a lot of people attacks settle once levels stop fluctuating. That is a pattern rather than a promise, and the timeline varies enough that any specific expectation will be wrong for someone.

What a year of monthly rates gives you is the ability to see which direction you are moving in, which is the only version of this question you can actually answer for yourself. “Worse than last year” and “worse than the worst year” are different situations, and neither is visible from inside a bad fortnight.

None of this is medical advice, and perimenopause is a period when new headaches deserve attention rather than assumption. A headache pattern that changes in character, comes on suddenly and severely, or arrives with new neurological symptoms needs a doctor rather than a diary — and after 50, a genuinely new headache is worth mentioning promptly rather than filing under hormones. The other tracking guides start from the same daily record.