Headaches that come from your neck — and how to know if yours do

Dr. Malek Sleem

Licensed Physical Therapist · FixUrMotion, Beirut

There is a particular kind of patient we see a lot of. They have had headaches for years. They have been to a GP, probably a neurologist. They have had an MRI or a CT, and it came back clean. They have a drawer with three different painkillers in it, and two of them used to work better than they do now. Nobody has ever put a hand on their neck.

That last part is the interesting one. The neck is one of the few structures capable of producing head pain that almost never gets examined during a headache work-up, because a headache is understood as a problem of the head. Sometimes it is not. This article is about how to tell.

How a neck problem becomes head pain

The mechanism is not vague, and it is worth understanding, because it is what makes the rest of this credible.

The sensory nerves from the top three segments of your neck — C1, C2 and C3 — feed into a relay station in the brainstem called the trigeminocervical nucleus. The trigeminal nerve, which supplies your face, forehead, temples and the area behind your eyes, feeds into the same place. Two different sets of inputs, one shared switchboard.

Your brain does not receive a return address. When a joint, a disc or a deep muscle in the upper neck sends a steady stream of irritation into that nucleus, the sensation can be interpreted as coming from the trigeminal side of the network — the eye, the temple, the forehead. This is called referred pain, and it is the same well-documented phenomenon that makes a heart attack hurt in the left arm. The pain is real, and it is genuinely felt where you feel it. The source is simply somewhere else.

When a headache is driven this way, it is called a cervicogenic headache — literally, born of the neck.

How common is it, honestly

You will find wildly different numbers quoted for this, and it is worth being straight about why. Population studies range from 0.17% in the Norwegian Akershus study to 4.1% in the Vågå study of headache epidemiology — a twenty-fold difference driven almost entirely by which diagnostic criteria the researchers applied. Among people with chronic headache, the figure most commonly cited is 15 to 20%.

So: a minority of all headaches, but a meaningful slice of the stubborn ones. That is the honest version, and it matters, because it means the neck is worth checking and is not the answer to everything.

The more striking number is about migraine. A 2022 systematic review and meta-analysis in Cephalalgia found neck pain in 77% of clinic-based migraine patients, rising to 87% in chronic migraine, and twelve times more prevalent than in people without headaches. Neck pain accompanied migraine more often than nausea did. Almost nobody would call migraine a neck condition on the back of that, and it would be wrong to — but it does suggest the neck is a participant in far more headaches than it usually gets asked about.

The regional picture explains the volume we see. Migraine prevalence across Arab countries has been reported anywhere from 2.6% to 32% depending on the population studied, and a survey of 1,144 Lebanese university students found 35.8% screening positive for migraine, with women at 42% against 23.3% of men. Meanwhile research on Saudi Ministry of Health office workers put twelve-month neck pain prevalence at 64%. Those two populations overlap far more than the separate specialties treating them do.

Telling the three apart

Most chronic headaches fall into three buckets, and they behave differently enough that you can usually get a strong hint from your own history before anyone examines you.

Two cautions on that table. First, the single most useful clue for the neck is side-locked pain that starts at the back and moves forward, combined with a neck that does not turn as far one way as the other. Second, these are not mutually exclusive. The Akershus study found 42% of its cervicogenic headache cases also had migraine. Plenty of people have two things at once, which is precisely why treating only one of them produces a partial result and a frustrated patient.

Why your scan was clean

This is the part that upsets people, and it should not. A clean MRI is good news that got delivered badly.

Imaging is ordered to rule out the things that would be catastrophic to miss: a tumour, a bleed, a malformation, inflammation. It is very good at that job, and a normal scan means those have been excluded. What it cannot do is show you a function problem. There is no MRI sequence for a stiff C1 to C2 segment, a deep neck flexor that fatigues after ninety seconds, or a joint that only starts referring after two hours at a laptop. Structure at rest is not the same as behaviour under load.

It cuts the other way too. Degenerative changes on a neck scan are so common in people with no pain at all that finding them tells you remarkably little on its own. Which is why cervicogenic headache is a clinical diagnosis — made by examining how your neck moves and whether that reproduces your headache, not by reading a picture.

The most useful single test we have is the cervical flexion-rotation test. The neck is fully flexed to lock out the lower segments, isolating rotation to C1 to C2, and rotation is measured each way. Normal is around 44 degrees per side. The test is considered positive at a total under 32 degrees, or a difference of 10 degrees or more between sides. In diagnostic accuracy studies it has reported sensitivity around 91% and specificity around 90% for C1 to C2 related headache. It takes about a minute, needs no equipment, and it is not part of a standard headache work-up anywhere.

The painkiller trap nobody warned you about

If you have been managing headaches with medication for years, this section matters more than the rest.

Medication-overuse headache is a recognised diagnosis, not a scolding. Under the international criteria it is headache on 15 or more days a month in someone with a pre-existing headache disorder, developing out of regular overuse of acute medication for more than three months. The overuse thresholds are lower than most people assume: 15 or more days a month for simple painkillers such as paracetamol or ibuprofen, and 10 or more days a month for triptans, opioids or combination analgesics.

More than 60 million people worldwide are thought to be affected, and it is now considered the most common secondary headache disorder. The trap is that it is invisible from the inside: the medication works, then wears off, then the pain returns slightly worse, so you take more. The pattern people describe is that the tablets used to fix it and now they only take the edge off.

Two of the numbers from earlier collide here. The Akershus study found 50% of its cervicogenic headache cases also had medication overuse. Someone can therefore be running an untreated neck problem and a medication-driven headache simultaneously, each making the other harder to see. Untangling that is a job for your doctor and a physiotherapist together, and it is not something to attempt by suddenly stopping your medication — with some drugs that is genuinely unsafe. Raise it with your prescriber first.

What the evidence says about treating the neck

For cervicogenic headache specifically, the research is unusually clear for a musculoskeletal condition:

• The landmark multicentre randomised controlled trial, published in Spine in 2002, put 200 people with cervicogenic headache into manipulative therapy, specific exercise, both, or a control group, and treated them for six weeks. Both manual therapy and specific exercise significantly reduced headache frequency and intensity, and the effects were still holding at the twelve-month follow-up. Combining them was not statistically superior to either alone, though about 10% more patients gained relief.
• A randomised trial of a six-week, twice-daily home programme — isometric craniocervical flexion, scapular retraction holds, isometric rotation and stretching — reported 76% of the exercise group achieving at least a 50% reduction in headache frequency, against 29% of controls.
• Training aimed specifically at the deep cervical flexors — the small stabilisers under the front of the neck, not the big surface muscles — outperformed general neck exercise. Which side of the neck you train appears to matter.

Notice what those studies have in common: six weeks minimum, daily work at home, targeted rather than general exercise, and a therapist progressing it. Nobody got better from a stretching video.

And the honest boundary: this evidence is for cervicogenic headache. Neck treatment is not a cure for migraine, which is a neurological condition needing medical management. What it can reasonably do for a migraine sufferer is take a genuine load off — in the many people carrying both, treating the neck removes one input feeding a sensitised system. That is worth having. It is not the same claim.

Three things you can test this week

• Keep a headache diary for fourteen days. Date, which side, where it started, what you were doing in the two hours before, and every dose of medication. Two weeks of this tells a clinician more than an MRI does, and the medication column often answers a question on its own.
• Check your rotation. Sitting tall, turn your head slowly right, then left, and compare how far each goes and where it pulls. A clear difference between sides in someone with one-sided headaches is a real finding, not a curiosity.
• Change one thing about the eight hours. Screen top at eye level, phone brought up to your face instead of your head going down to it, and stand up every 40 minutes. Not because posture is a moral failing, but because sustained load is what keeps these tissues irritated, and frequency of movement beats duration of stretching.

If your headaches are largely one-sided, start in the neck, and your rotation is asymmetric, get the neck assessed properly — that combination is worth one short call before you spend another year on the same three tablets. Our guide to TMJ and jaw pain covers the jaw side of the same system, which is often involved when the headaches sit around the temples, and if getting to a clinic is the obstacle, supervised online rehab is a legitimate route for this kind of problem.

Frequently asked questions

Dr. Malek Sleem

Licensed physical therapist and founder of FixUrMotion in Beirut, Lebanon. Helping people move better, recover faster, and live pain-free, in-clinic and at home.

This article is general information, not a diagnosis. Headaches have many causes and a proper assessment is individual. If you are on regular headache medication, speak to your prescriber before changing anything.

Has anyone actually checked your neck?

If your headaches are one-sided, start at the back of your head, and nobody has ever examined the neck underneath them — that is worth one short call. We'll tell you honestly whether this is something we can help with.