The Cannabis Question: Relief, Trigger, or Both?

In 1892, Sir William Osler — a founder of modern medicine — called cannabis “probably the most satisfactory remedy” for migraine. Doctors prescribed it for the next 45 years. Then prohibition arrived, research stopped, and a century of silence followed.

That silence just broke.

In 2024, researchers at the University of California, San Diego published results from the first placebo-controlled trial of cannabis for acute migraine — 92 participants, four treatments, double-blind crossover design.^[1] The findings cracked open a door that mainstream neurology had kept locked for decades.

But the cannabis-migraine story doesn’t end at “it works.” The same plant that can abort an attack may also be priming you for more of them.

Welcome to the most complicated question in migraine medicine.

Your Brain Already Makes Cannabis

Before you touch a single leaf, your body runs its own cannabis-like system. The endocannabinoid system (ECS) is a network of receptors, enzymes, and signaling molecules that helps regulate pain, inflammation, mood, and nausea.^[2] Its two lead molecules — anandamide (AEA) and 2-arachidonoylglycerol (2-AG) — act like dimmer switches on pain signals. Turn them up, and the volume drops.

Here’s where migraine enters the picture.

In 2007, researchers at the University of Perugia measured anandamide in the cerebrospinal fluid — the liquid that bathes the brain — of people with chronic migraine. The gap stunned them: migraine patients had dramatically lower anandamide levels than healthy controls, with a p-value below 0.0001.^[3] That’s the kind of statistical confidence researchers almost never see.

This finding became the cornerstone of a hypothesis first proposed by neurologist Ethan Russo in 2004: Clinical Endocannabinoid Deficiency. The idea is elegant. Migraine, fibromyalgia, and irritable bowel syndrome — three conditions that frequently overlap and resist standard treatment — might share a single root: an endocannabinoid system running on empty.^{[4, 5]}

If true, supplementing that system with plant-derived cannabinoids isn’t masking symptoms. It’s replacing something the body fails to produce in sufficient quantities.

An elegant theory. But until recently, it lacked a critical piece: a proper trial.

The Trial That Changed the Conversation

The UC San Diego trial, led by headache neurologist Nathaniel Schuster, gave each participant four treatments for four separate migraine attacks: vaporized 6% THC, 11% CBD, a combination of 6% THC plus 11% CBD, or placebo cannabis flower with the active compounds chemically extracted. Nobody — not patients, not researchers — knew which was which.^{[1, 6]}

At two hours, the THC-CBD combination told a clear story. Pain relief: 67.2% versus 46.6% for placebo. Complete pain freedom — the toughest endpoint in migraine research — reached 34.5% compared to 15.5%. Freedom from the most bothersome symptom: 60.3% versus 34.5%. Benefits held at 24 and 48 hours.^[1]

Two findings forced a double-take.

First, THC alone reduced pain (68.9% vs. 46.6%) but failed to achieve pain freedom or most bothersome symptom freedom. CBD alone? It performed no better than placebo on any measure.^{[1, 6]} This matters enormously. The CBD industry has marketed cannabidiol as a migraine remedy for years. This trial says otherwise — at least for acute attacks. It’s the combination that works.

Second — and this is the hidden gem of the data — the THC-CBD combination specifically reduced photophobia and phonophobia but not nausea.^[1] Why does that matter? THC is a well-known anti-nausea agent. If the benefit came simply from the “high” dampening symptom awareness, nausea should improve first. Instead, the combination targeted the exact sensory processing abnormalities that define migraine as a neurological event. Not just pain. The disease itself.

“This is the first real — to me — compelling evidence for the anti-migraine effects of cannabis in humans,” Schuster said when presenting final results at the 2025 American Headache Society meeting.^[6] He also emphasized a point that deserves repeating: patients who treated earlier — within the first two hours of an attack — responded better than those who waited. Early treatment wins. Always.

The question the trial couldn’t answer is why the combination worked — and that answer lives in a molecule you may already know.

The CGRP Connection Hiding in Plain Sight

If you’ve followed migraine science, you know CGRP — calcitonin gene-related peptide — the molecule behind the newest migraine drugs, from Aimovig to Nurtec. When the trigeminal nerve fires during an attack, it floods the meninges with CGRP, igniting inflammation and pain.^[7]

Cannabinoids intersect this pathway in ways that startled researchers.

Anandamide — the same molecule found depleted in migraine patients’ spinal fluid — directly inhibits CGRP release from trigeminal nerve fibers through CB1 receptor activation.^{[7, 8]} In animal models, a synthetic anandamide analog reduced CGRP levels in plasma, trigeminal ganglia, and the brainstem while blocking mast cell degranulation — one of the key inflammatory events in migraine.^[9]

A 2024 review in The Journal of Headache and Pain mapped six immune cell types that carry both cannabinoid receptors and CGRP receptors: macrophages, monocytes, mast cells, dendritic cells, B cells, and T cells.^[7] The implication is provocative: cannabinoids could theoretically modulate the same pathway targeted by existing CGRP-blocking drugs. Nobody has tested this combination clinically. But the molecular framework is there, and research teams from Iowa to Brazil are building the case.

In early 2025, a preclinical study in Cephalalgia confirmed the pattern across three different mouse migraine models and two independent laboratories. Photophobia improved only with a CBD-THC combination — not with either compound alone.^[10] The mirror image of the human trial. Two species. Same answer.

The Dark Side: When Relief Becomes Rebound

Now for the part nobody wants to hear.

In 2021, Stanford University researchers reviewed records of 368 chronic migraine patients. Among cannabis users, 81% had medication overuse headache — the dreaded “rebound.” Among non-users, the rate was 41%. After adjusting for every available confounder, cannabis users were six times more likely to suffer rebound headache.^[11]

Six times. That number demands attention.

The researchers also uncovered a bidirectional link between cannabis and opioid use: people using one were significantly more likely to use the other.^[11] Both substances influence the periaqueductal gray — a brainstem region that acts as a master volume knob for pain processing — and both can recalibrate the brain’s pain threshold when used too frequently.^{[11, 12]}

The paradox cuts deep. The very endocannabinoid system running low in migraine can become further dysregulated by too-frequent external cannabinoid exposure. It’s structurally identical to the triptan paradox — the most effective acute treatment becomes the engine of chronic headache when overused.

A separate analysis of over 12,000 cannabis-treated headache sessions found that while acute severity dropped by roughly 50%, doses escalated over time — users needed progressively more to achieve the same effect.^[13] For general headache, efficacy also declined with repeated use. For migraine specifically, efficacy appeared more stable, though consumption still crept upward.^[13] Tolerance in one domain. Not the other. Nobody fully understands why.

The lead researcher of the Stanford study was candid about its limits: because the design was retrospective, the direction of causality remains unclear. Were patients using cannabis because they already had rebound headache? Or did cannabis contribute to developing it? The data can’t separate the two.^[11] But the signal is strong enough to warrant caution.

The Estrogen Factor

Migraine affects three times more women than men. The endocannabinoid story has a gender chapter — and it may be the most underappreciated angle in this entire field.

Women with migraine show heightened activity of FAAH, the enzyme that breaks down anandamide, along with increased activity of the anandamide membrane transporter in their platelets.^{[14, 15]} The net effect: faster anandamide destruction, lower levels.

The driver? Estrogen. It regulates FAAH expression. When estrogen drops — during menstruation, perimenopause, or postmenopause — FAAH activity rises, anandamide plummets, and the pain threshold falls.^{[15, 16]} This may partly explain why menstrual migraine is so notoriously difficult to treat, and why some women report that cannabis provides specific relief during their period.

Preclinical research consistently finds female rodents more sensitive to THC’s effects than males.^[16] At the University of Arizona, researcher Tally Largent-Milnes has been mapping endocannabinoid system expression in brain regions tied to migraine — including the periaqueductal gray — and finding it fundamentally different between sexes.

“This seems to be a female-mediated system,” Largent-Milnes said. “It’s likely that hormone regulation of the endocannabinoid system is driving female susceptibility to headache.”^[17]

Her lab is now investigating whether boosting 2-AG — the second major endocannabinoid, which has been largely overlooked in migraine research — could offer targeted relief without the side effects of plant cannabis.^[17] It’s early. But it points toward a future where the treatment isn’t the plant itself, but drugs designed to enhance what your body already makes.

What This Means for You

The science draws clear lines in some places — and honest blanks in others.

What we know: A specific combination of low-dose THC and CBD, delivered by vaporization early in a migraine attack, outperforms placebo in the first rigorous trial of its kind.^[1] CBD alone doesn’t work for acute attacks.^[1] The endocannabinoid system is measurably disrupted in chronic migraine.^{[3, 4, 5]} Cannabis interacts with the same CGRP pathway targeted by the newest migraine medications.^{[7, 8]} And frequent use carries a real, quantifiable risk of medication overuse headache.^{[11, 12]}

What we don’t know: The optimal dose for prevention. Whether cannabis can safely combine with CGRP-targeting drugs. How long-term use affects migraine frequency over years. Whether edibles or tinctures offer a safer profile than inhalation. Whether certain genetic profiles predict who benefits and who doesn’t.

Schuster himself put it plainly: standard treatments — triptans, gepants — should be tried first. Cannabis is a conversation to have with your doctor when conventional options haven’t worked, not a first-line choice.^[6]

After 30 years of living with migraine, what strikes me most is that we’re finally having this conversation with data instead of anecdotes. For a century, patients who found relief with cannabis had nothing but their own experience as evidence. Now there’s a randomized controlled trial in Headache, a landmark CSF study in Neuropsychopharmacology, and a growing body of preclinical work connecting cannabinoids to the molecular architecture of migraine itself.

The question was never really whether cannabis “works.” It was whether we’d take it seriously enough to find out how, for whom, and at what cost.