Migraine Came Back Hours Later? Why Triptans Wear Off — and What to Change

A woman in her late thirties sat down in my office last month and said something I hear almost every week: “The sumatriptan works. That’s the problem.” She meant that two hours after the tablet, the pain was gone — and by dinner it was back, the same headache, the same side. Over a bad weekend she would take three or four tablets chasing it.

What she was describing has a name: headache recurrence. Roughly one in three migraine attacks treated with a triptan comes back within 24 hours, and it is a fundamentally different problem from a triptan that never worked at all. Recurrence usually means the attack simply outlasted the drug — not that the drug is wrong for you. In most cases it can be fixed by treating earlier, changing how the drug gets into you, pairing it with a long-acting anti-inflammatory, or moving to a longer-acting agent. The one thing you should not do is keep redosing your way through it, because that is the road to medication-overuse headache.

Recurrence and non-response are two different failures

This distinction matters more than almost anything else in the conversation, and it is the first thing I ask about. If you take a triptan and two hours later the pain is unchanged, that is non-response — the drug did not engage the attack, and a second dose of the same drug is very unlikely to help. If you take a triptan and get real relief, then the headache creeps back four, eight, or twelve hours later, that is recurrence — and a second dose usually does work, because the drug clearly has an effect on this attack.

Patients often lump both under “my triptan doesn’t work,” and so do some prescribers, which leads to the wrong fix. If your problem is genuine non-response — nothing happens, ever — I have written separately about why triptans fail and what a neurologist tries next. The rest of this article is about the other pattern: it works, and then it comes back.

Why the attack outlasts the drug

An untreated migraine attack runs anywhere from four to seventy-two hours. Most triptans clear the body much faster than that. Sumatriptan has a half-life of about two hours; rizatriptan and zolmitriptan are in the same range; eletriptan sits at roughly four. If your natural attack is a thirty-six-hour event and your drug is gone in eight, the biology underneath was never finished — it was suppressed, and it resurfaced. That is the single commonest reason for recurrence, and it explains why people with long attacks, and women with menstrual migraine in particular, run into it more often than anyone else.

Timing is the second reason. Triptans work best taken early, while the pain is still mild and while it is still confined to the head. Once allodynia has set in — that state where brushing your hair, wearing glasses, or a shirt collar against your neck hurts — central pain pathways are already sensitized, and a triptan given at that point is both less likely to abolish the pain and more likely to be followed by recurrence. Many of my patients wait deliberately: they want to be sure it is “a real one” before spending a tablet. That instinct is understandable and it is costing them.

The third reason is absorption. Migraine slows the stomach — gastroparesis is part of the attack itself, not just a side effect of nausea — so an oral tablet swallowed during a vomiting-grade attack may be sitting undigested when you most need it in your bloodstream. A drug that arrives late and incompletely produces exactly the picture of partial relief followed by return of pain.

What I change first, before changing the drug

I start with the cheapest interventions, because they solve a surprising share of cases.

Treat earlier. Take the triptan while the pain is mild and you can still work — not at the point where you are lying down in the dark. If you get an aura, current guidance is to take the triptan when the head pain begins rather than during the aura itself.

Change the route. If you are nauseated or vomiting early, an oral tablet is the wrong delivery system. Subcutaneous sumatriptan and nasal formulations bypass the gut entirely and reach effective levels within minutes. An orally disintegrating tablet is easier to keep down but still has to be absorbed through the stomach, so it does not solve the gastroparesis problem — a point that surprises most people. Adding an antiemetic such as metoclopramide can improve both the nausea and the absorption.

Add a long-acting NSAID. This is my highest-yield move for recurrence. Naproxen has a half-life of twelve to seventeen hours — it is still working long after the triptan has cleared. Taken together with the triptan at the start of the attack, it measurably reduces the odds of the headache returning between two and twenty-four hours; the fixed-dose sumatriptan–naproxen combination was developed on exactly this rationale. It is not for everyone: if you have ulcer disease, kidney impairment, are on an anticoagulant, or are pregnant, this needs a conversation with your doctor first.

Use the second dose properly. For recurrence, a repeat dose after the interval on your label is reasonable and usually effective, within the daily maximum. What I want you to notice is the pattern, not the individual day. Two tablets to cover one long attack is not a crisis. Two tablets every attack, several attacks a week, is a system that has already failed.

When the answer really is a different drug

If the basics do not hold the attack, pharmacokinetics become the argument. Naratriptan and frovatriptan are the long-acting members of the family — frovatriptan’s half-life is around twenty-six hours — and in trials they show somewhat lower recurrence rates. The trade-off is real: they are slower and gentler in onset, so people who value fast knockdown often dislike them. For women whose recurrence is concentrated around menstruation, that long half-life can be used deliberately as short-term prevention across the vulnerable days, rather than as rescue.

The gepants — ubrogepant and rimegepant — are the other option, and their longer half-lives make them a rational choice when recurrence is the dominant pattern. They also carry no vasoconstrictive risk, which matters if cardiovascular disease limited your triptan choices in the first place. So far they have not shown the medication-overuse pattern that triptans and analgesics do, though they are newer drugs and I stay conservative about how confidently I say that. I covered how this class works in more detail in my guide to CGRP-targeted migraine treatment.

One caution before you chase a better rescue drug: if an attack has now run past seventy-two hours without a meaningful break, that is status migrainosus, and it needs to be broken medically rather than managed with repeated triptan doses at home.

The number that decides everything

Here is the part I care most about. Count the number of days per month you take anything for headache — triptans, gepants, combination painkillers, plain ibuprofen, everything. If that number is reaching ten days a month for triptans or combination analgesics, or fifteen for simple painkillers, and has been for three months, you are in the territory where the treatment itself starts driving the headaches. Recurrence is one of the most common on-ramps to that problem, precisely because it feels so justified in the moment: the pain is real, the drug helped before, so you take another one.

And there is a strategic point buried in the counting. If you are treating on ten days a month, you do not primarily have an acute treatment problem — you have an attack frequency problem, and the honest fix is preventive treatment, not a cleverer rescue plan. Recurrence forces that question into the open, which is one of the few good things about it. If you have been cycling through acute drugs for a year without anyone asking how many days a month you are treating, it is worth reading why headaches so often fail to get better — the answer is usually a diagnostic or strategic gap, not a missing prescription.

The woman I mentioned at the start turned out to have twelve headache days a month, not the “three or four bad ones” she had reported. She had been counting attacks, not days. We moved her triptan earlier, added naproxen alongside it, and started a preventive. The recurrence did not need a heroic drug. It needed someone to count properly.

---

Next steps

Before your next appointment, check your symptoms against my free headache red-flag checklist — it takes two minutes and tells you which features need urgent assessment rather than a better tablet.

If your headaches keep coming back and no one has stepped back to look at the whole picture, you can talk to a specialist here.

Dr. Tae-Ho Guak, MD, is a board-certified neurologist with 25 years of practice and over 100,000 patients seen.

This article is for education only and does not replace an individual medical consultation. Do not start, stop, or change any medication without speaking to your own doctor. If you have a sudden, severe headache that peaks within seconds, a headache with fever and neck stiffness, weakness, confusion, or a first seizure, go to an emergency department immediately.