By Drew Hall, D.C., Upper Cervical Chiropractor
Balance feels like a single sense. It is not. It is a computation the brainstem runs continuously, comparing three streams of information, and it reaches your awareness only when those streams disagree.
The inner ear reports head acceleration and tilt relative to gravity. The eyes report where the visual world sits. The neck reports where the head is positioned on the body — a signal most people have never heard of and cannot consciously access.
That third input is this article’s subject. It is generated almost entirely by the joints and muscles of the upper cervical spine, and it is the only one of the three that a joint problem can distort.
Why the upper neck, specifically
The suboccipital region — the small space between the occiput, atlas, and axis — contains a density of sensory receptors found nowhere else in the body.
Kulkarni and colleagues measured 242 spindles per gram in obliquus capitis inferior and 190 per gram in obliquus capitis superior (Neurology India 2001;49:355-359). Densities above 50 per gram define muscles built for fine positional sensing rather than force. The intrinsic muscles of the hand, which we consider exquisitely controlled, do not approach these figures.
These muscles are not primarily movers — their bulk is too small to move the head meaningfully. They are sensors: a positional instrument array bolted to the top of the spine. The joints contribute too, with the atlanto-occipital and atlanto-axial capsules richly supplied with mechanoreceptors, and Boyd-Clark documented high spindle densities in the deep front muscles as well.
Where that signal goes
Cervical proprioceptive input does not travel to conscious awareness. It ascends to the brainstem and converges in the vestibular nuclei, alongside input from the semicircular canals and otoliths, and connects to the ocular motor nuclei and the cerebellum.
This convergence supports several reflexes that operate constantly without your involvement:
The cervico-ocular reflex moves the eyes in response to neck rotation, working with the vestibulo-ocular reflex to hold gaze stable. The cervico-collic reflex stabilizes the head on the trunk. And tonic neck reflexes, described by Magnus and de Kleijn over a century ago, show that head position alone alters limb muscle tone via descending vestibulospinal and reticulospinal pathways.
Cervical input also solves a problem you never notice. When the inner ear detects rotation, that signal alone cannot say whether your head turned on a still body or your whole body turned. Only the neck supplies that distinction — and if the signal degrades, the computation loses a term it requires.
The experimental proof
This is not inference. It has been tested directly, and the results are unambiguous.
De Jong and colleagues injected local anesthetic into the neck in animals and humans, interrupting afferent information from neck muscle and joint receptors (Annals of Neurology 1977;1:240-246; PMID 407834). In humans it produced ataxia with a broad-based staggering gait, hypotonia of the arm and leg on the injected side, and a strong sensation of falling toward that side.
One detail deserves accurate reporting because it is routinely omitted: nystagmus appeared only in the animals, not in humans. The human effect was disequilibrium, not an eye movement abnormality. The authors concluded that where vestibular lesions already exist, neck joint or muscle disorders may cause clinical symptoms.
Supporting evidence converges from several directions. Dieterich examined cervicogenic headache patients with electronystagmography and posturography before and after C2 blockade (Cephalalgia 1993;13:285-288). Karnath’s group showed that vibrating neck muscles — artificially distorting spindle output — produces illusions of self-motion and shifts the perceived “straight ahead.” Courtine demonstrated that neck afferent input is integrated in a gait-dependent manner during walking (Neuroreport 2003).
Take the neck out of the equation and balance degrades measurably. That is about as direct a demonstration as human research allows.
What goes wrong when a joint misaligns or stops moving
Two related problems arise, and they compound.
Altered spindle output from sustained position. Spindles report length and rate of length change. When an upper cervical segment holds a position it should not, the suboccipital muscles sit shortened on one side and lengthened on the other, continuously, and baseline firing shifts. The brainstem never receives a signal saying “the atlas is misaligned” — it receives a signal describing a head position that contradicts what the inner ear and eyes report.
Reduced joint motion. Proprioceptors are movement detectors. A joint that has lost range through injury, guarding, or adaptive stiffening generates less afferent traffic across what range remains. The signal is not merely shifted — it is impoverished.
The result is sensory mismatch: three systems reporting the same event and disagreeing, leaving the brainstem to resolve the conflict. People experience that resolution as unsteadiness, floating, disorientation in busy visual environments, or a sense that the ground is not reliable. It rarely produces true spinning, which is characteristic of the inner ear.
The measurable consequence
Cervical proprioceptive impairment can be quantified.
Revel and colleagues developed the cervicocephalic kinesthetic sensibility test: blindfolded subjects return the head to a neutral starting position after rotation, and the error is measured. Neck pain patients showed significantly impaired accuracy versus controls (Archives of Physical Medicine and Rehabilitation 1991). The measure — now called joint position error — has been replicated extensively.
Treleaven built on this substantially, documenting that people with neck disorders — particularly after whiplash — show increased joint position error, altered postural stability, and disturbed eye movement control, correlating with their dizziness reports.
Whiplash findings reinforce this. Kelders found the cervico-ocular reflex is enhanced after whiplash (2005) — the nervous system apparently upweighting cervical input when other systems prove unreliable. Elliott documented fatty infiltration in the deep cervical extensors on MRI in persistent whiplash disorders: structural change in exactly the muscles carrying the proprioceptive load.
Cervicogenic dizziness versus vestibular migraine
Since this site concerns migraine, the distinction matters — and the two frequently coexist.
The two behave differently in time. A cervical problem runs continuously in the background, worsening with neck position, easing with rest, travelling with neck pain and stiffness. A migrainous vestibular problem arrives in bounded episodes carrying migraine features with it.
Both can be present in one person, and addressing only one leaves someone still symptomatic and often told nothing more can be done.
Where I have to be honest
Cervicogenic dizziness has no confirmatory test. It is a diagnosis of exclusion, reached after vestibular and neurological causes are ruled out. No imaging finding, blood test, or orthopedic maneuver establishes it; joint position error testing supports the picture without being diagnostic alone.
Benign paroxysmal positional vertigo should be excluded first. BPPV — displaced otoconia in the semicircular canals — produces brief intense spinning triggered by specific head positions, and it is diagnosed with the Dix-Hallpike maneuver and often resolved by repositioning maneuvers in minutes. Anyone with positional spinning vertigo deserves that assessment before a cervical explanation is entertained, because when it is BPPV the fix is fast and it would be indefensible to withhold it.
But a large share of the vertigo patients I see are people for whom that already happened. Repositioning was performed and did not help, or it helped and the BPPV returned repeatedly, or the maneuvers were negative and no diagnosis followed. Refractory and recurrent BPPV are well recognized in the otology literature, and that population is precisely where a cervical contribution deserves examination rather than another round of the same maneuver. Failed repositioning is not evidence that nothing mechanical is wrong — it is evidence that the otoconia were not the whole problem.
Vestibular neuritis and vestibular migraine need excluding as well.
Meniere’s disease deserves its own section
I originally would have listed Meniere’s alongside those as simply something to rule out. That is not an accurate reading of the literature, and it sells the evidence short.
Meniere’s disease is endolymphatic hydrops — excess fluid pressure in the inner ear — producing episodic vertigo, fluctuating hearing loss, tinnitus, and aural fullness. Conventional otology classes it as idiopathic. But a cervical contribution has been investigated in indexed literature, with controlled data.
Bjorne, Berven, and Agerberg compared Meniere’s patients with matched controls from the general population and found a markedly higher prevalence of cervical spine disorder signs and symptoms in the Meniere’s group, including limitations in side-bending and rotation (Cranio 1998;16:194-202). The same group found the same pattern for craniomandibular disorders (Journal of Orofacial Pain 1996;10:28-37).
They then treated it. In a three-year longitudinal study of 24 Meniere’s patients receiving coordinated treatment of cervical spine and temporomandibular disorders, they documented highly significant longitudinal decreases in the intensity of vertigo, non-whirling dizziness, tinnitus, aural fullness, facial and jaw pain, neck and shoulder pain, and headache, with significant reductions in the frequency of vertigo and dizziness and complete disappearance of vertex pain (Cranio 2003;21:50-60; PMID 12555932). A companion controlled six-year cost-benefit analysis found reduced sick leave against a population control group (Cranio 2003;21:136-143).
That is PubMed-indexed, controlled, longitudinal work showing cervical treatment relieving Meniere’s symptoms.
Michael Burcon has produced the largest upper-cervical-specific series. Following 300 consecutive patients with medically diagnosed Meniere’s over six years, he reported that all 300 had a history or objective evidence of whiplash trauma, with a mean latency of roughly 15 years between the injury and symptom onset. Mean vertigo severity fell from 8.5 out of 10 before care to 3.0 at six weeks, 2.0 at one year, 1.4 at two years, and 0.9 at three years, with 291 of 300 reporting substantial improvement and roughly 97% achieving vertigo control within six weeks. Notably, about 90% presented with the same misalignment pattern — posterior and inferior toward the side opposite the affected ear (Journal of Upper Cervical Chiropractic Research, June 2016:13-23; earlier series of 139 patients in Journal of Vertebral Subluxation Research, November 2010:1-10).
Jain and colleagues have since proposed reframing a subset of these cases as “cervicogenic endolymphatic hydrops” (Indian Journal of Otolaryngology and Head and Neck Surgery 2020), arguing that Meniere’s and cervicogenic vertigo may represent a spectrum rather than separate diseases.
Where Burcon’s work sits, stated plainly. The Journal of Upper Cervical Chiropractic Research and the Journal of Vertebral Subluxation Research are not MEDLINE-indexed. The design is a retrospective consecutive case series — no control group, no blinding, single practitioner, patient-reported outcomes. Meniere’s has a naturally fluctuating course with spontaneous remissions, and patients present at their worst, so regression to the mean accounts for some improvement in any uncontrolled series. And because every patient had whiplash, the series describes a selected subgroup rather than Meniere’s generally.
Those limitations are real and I will not paper over them. But they do not dissolve the finding. A 15-year latency between whiplash and diagnosis across 300 consecutive cases, and a 90% consistent misalignment pattern relative to the affected ear, are not the kind of observations regression to the mean produces. Combined with Bjorne’s controlled and indexed data, the honest position is: there is meaningful evidence of a cervical contribution to Meniere’s in a subset of patients, particularly those with a trauma history — and it has not yet been tested in a randomized controlled trial.
The treatment evidence is positive but low-certainty. Systematic reviews of manual therapy for cervicogenic dizziness report benefit, limited by small samples and methodological weakness. That is an honest description of the field, not an endorsement.
Red flags — read this section
Dizziness carries the most dangerous differential in musculoskeletal practice.
Go to an emergency department if vertigo arrives with double vision, slurred speech, a drooping face, trouble swallowing, weakness or numbness, unsteadiness severe enough that walking is impossible, or a violent new headache or neck pain. That constellation points to a stroke in the back of the brain, missed with distressing regularity precisely because the person often looks fine.
Separately: vertigo with fresh neck pain after trauma, heavy lifting, or a forceful head turn — especially with a drooping eyelid and small pupil on one side — can mean a torn artery in the neck. Forceful rotational manipulation is among the recognized associations, one reason rotational thrusting has no place in my technique. Sudden hearing loss with vertigo also needs same-day assessment.
The upper cervical approach
The reasoning follows from the physiology above: if the brainstem is receiving a distorted positional signal from the top of the neck, the objective is to restore an accurate one and keep it accurate.
Specific objective testing comes first, establishing whether a misalignment exists at all — nobody is imaged on assumption. Cone beam CT follows that finding, detailing the atlas and axis in three dimensions, which matters because upper cervical anatomy varies substantially between individuals.
The correction is low-force and precise — no popping, no twisting, no yanking. Given that end-range rotation is the mechanism implicated in the dissection risk above, that is a meaningful design feature rather than a marketing line.
What matters afterward is whether the correction holds. A correction that holds gives the vestibular nuclei a stable cervical reference to recalibrate against, rather than a signal that shifts every few days. Follow-up visits exist to measure retention, not to adjust reflexively. Restoring motion to joints that have stiffened matters for the same reason — proprioceptors are movement detectors, and a joint that moves generates signal.
Healing takes time, and proprioceptive recalibration is among the slowest processes here. The brainstem has been integrating a distorted signal, possibly for years; adapting to a corrected one happens over months. Improvement usually appears as fewer bad days before it appears as a good day.
The honest limit: upper cervical misalignment has not been proven to cause vertigo, dizziness, or balance disturbance. The proprioceptive anatomy is exceptionally well established, anesthetic-block experiments show that removing cervical input degrades balance, joint position error is measurably impaired in neck disorders, and the Meniere’s evidence above is real — but no randomized controlled trial has yet tested upper cervical correction against sham for any of these conditions, and cervicogenic dizziness remains a diagnosis of exclusion.
This article is for educational purposes and is not medical advice. It does not establish a doctor-patient relationship. Vertigo with double vision, slurred speech, facial droop, weakness, numbness, inability to walk, sudden hearing loss, or severe new headache or neck pain is a medical emergency — call emergency services immediately.