Treatment in cervicogenic dizziness is mechanism-specific, not diagnosis-specific. Two patients carrying the same label can need materially different care depending on which of the four mechanisms from Module 3 is doing the work in their presentation. The recurring failure mode is to treat the label rather than the mechanism — and the recurring rescue is to reconsider which mechanism is actually in play when the first attempt does not progress.
The short version
Once a cervicogenic diagnosis has been made — which means the vestibular workup is clean and the other diagnoses have been excluded (Modules 5 and 7) — treatment goes in one of four directions depending on which underlying mechanism dominates:
Proprioceptive mismatch (the typical patient) — manual therapy and cervical proprioceptive retraining. Most patients fit here.
Vestibulo-sympathetic loop — the same as above, plus targeted help for the autonomic features (nausea, palpitations).
Rotational vertebral artery syndrome — avoidance of provocative positions, vascular risk management, and surgical referral when a structural cause is confirmed. No manipulation.
Central reweighting / chronic — vestibular rehabilitation, visual desensitisation, and addressing the avoidance behaviour that perpetuates the condition.
The tool below lets you pick a mechanism and see what to do, what to add, and importantly, what to avoid in that route.
Mechanism-routed treatment selector
The four mechanisms from Module 3 require different treatments. Pick the mechanism that dominates this patient's presentation and the pathway below updates — first-line interventions, adjuncts, and (importantly) what to avoid.
Proprioceptive mismatch
M3 §Mechanism 1
Disordered cervical afferent input. The typical mechanically-presenting patient with neck pain, restricted segmental motion, and abnormal SPNT or JPE.
First-line
Upper-cervical spinal mobilisation and manipulation. Moderate evidence (level 2) supports this as a first-line route, with consistent improvement in dizziness severity and symptom frequency across RCTs.[lystad-2011]
Cervical proprioceptive retraining — joint position retraining drills using a head-mounted laser pointer (the protocol Revel et al validated).[revel-1991]
Specific deep cervical flexor strengthening and craniocervical motor-control work, building tolerance to sustained head positions.[jull-2008]
Adjuncts
Smooth pursuit neck torsion retraining — coupling gaze-stabilisation drills with controlled torsion, building gain in the impaired position.[treleaven-2017]
Postural and balance training that progressively challenges the cervical-proprioceptive channel.[sremakaew-2023]
Pain modulation as needed — short-course NSAIDs, heat, paced activity. Pain itself disturbs cervical spindle gain, so reducing pain reduces mismatch.
Avoid
Pure vestibular habituation drills as a primary intervention — they treat a peripheral lesion that this patient does not have, and may reinforce the visual-dependence reweighting (see Mechanism 4).
Long-term cervical collar immobilisation — it deconditions the deep flexors and reduces normal afferent flow, both of which worsen the underlying disturbance.
Vestibular suppressants (cinnarizine, prochlorperazine, betahistine for non-Ménière indications) beyond brief symptomatic cover. They blunt central compensation.
Expected timeline
Most published trials show medium-to-large effects at 6–12 weeks. Adding sensorimotor training to manual therapy alone produces gains that are better maintained at 12 months than manual therapy in isolation.
Figure 8.1 — Mechanism-routed treatment. The same symptom — chronic dizziness with neck pain — calls for materially different interventions depending on which of the four mechanisms dominates. Manual therapy in a centrally-reweighted patient produces months of effort without progress; vestibular habituation in undiagnosed RVAS courts harm; identical-looking patients require non-identical care.
One sentence to remember. The same diagnosis on two patients can call for opposite treatments — manual therapy helps the mechanical patient and stalls the centrally-reweighted one; vestibular habituation helps the chronic patient and worsens the RVAS patient. The mechanism is the prescription.
The four-route framework
The four mechanisms from Module 3 — proprioceptive mismatch, the vestibulo-sympathetic loop, rotational vertebral artery syndrome, and central reweighting — are not mutually exclusive in any one patient, but one usually dominates the clinical picture. The treatment plan should be built around the dominant mechanism, with adjuncts where secondary mechanisms are contributing.
Mechanism-routed treatment selector
The four mechanisms from Module 3 require different treatments. Pick the mechanism that dominates this patient's presentation and the pathway below updates — first-line interventions, adjuncts, and (importantly) what to avoid.
Proprioceptive mismatch
M3 §Mechanism 1
Disordered cervical afferent input. The typical mechanically-presenting patient with neck pain, restricted segmental motion, and abnormal SPNT or JPE.
First-line
Upper-cervical spinal mobilisation and manipulation. Moderate evidence (level 2) supports this as a first-line route, with consistent improvement in dizziness severity and symptom frequency across RCTs.[lystad-2011]
Cervical proprioceptive retraining — joint position retraining drills using a head-mounted laser pointer (the protocol Revel et al validated).[revel-1991]
Specific deep cervical flexor strengthening and craniocervical motor-control work, building tolerance to sustained head positions.[jull-2008]
Adjuncts
Smooth pursuit neck torsion retraining — coupling gaze-stabilisation drills with controlled torsion, building gain in the impaired position.[treleaven-2017]
Postural and balance training that progressively challenges the cervical-proprioceptive channel.[sremakaew-2023]
Pain modulation as needed — short-course NSAIDs, heat, paced activity. Pain itself disturbs cervical spindle gain, so reducing pain reduces mismatch.
Avoid
Pure vestibular habituation drills as a primary intervention — they treat a peripheral lesion that this patient does not have, and may reinforce the visual-dependence reweighting (see Mechanism 4).
Long-term cervical collar immobilisation — it deconditions the deep flexors and reduces normal afferent flow, both of which worsen the underlying disturbance.
Vestibular suppressants (cinnarizine, prochlorperazine, betahistine for non-Ménière indications) beyond brief symptomatic cover. They blunt central compensation.
Expected timeline
Most published trials show medium-to-large effects at 6–12 weeks. Adding sensorimotor training to manual therapy alone produces gains that are better maintained at 12 months than manual therapy in isolation.
Figure 8.1 — Mechanism-routed treatment. The same symptom — chronic dizziness with neck pain — calls for materially different interventions depending on which of the four mechanisms dominates. Manual therapy in a centrally-reweighted patient produces months of effort without progress; vestibular habituation in undiagnosed RVAS courts harm; identical-looking patients require non-identical care.
The evidence base, in brief
Manual therapy
The most-studied intervention in cervicogenic dizziness. The Reid & Rivett 2005 systematic review found Level 3 evidence for manual therapy across nine studies of mixed methodological quality.11 Lystad et al's 2011 systematic review of fifteen articles, including five RCTs, concluded that there is moderate evidence (Level 2) in favour of manual therapy — particularly cervical spinal mobilisation and manipulation — for cervicogenic dizziness, with consistent improvement in dizziness severity and frequency, neck pain, range of motion, postural stability, and joint position sense.42 Reid et al's 2014 RCT of cervical manual therapy demonstrated improvements in dizziness symptoms, head repositioning, and balance at 12 weeks post-intervention with effects maintained at follow-up.43
Sensorimotor training and the Treleaven protocol
Adding cervical sensorimotor training (joint position retraining, smooth pursuit neck torsion retraining, dynamic balance with controlled cervical input) to manual therapy produces gains that are better maintained at 12 months than manual therapy alone. The Sremakaew et al 2023 RCT randomised 152 participants with neck pain to four intervention arms and demonstrated medium-to-large additional benefit from sensorimotor training, maintained at 12-month follow-up.44 This is the highest-quality evidence supporting the routine combination of manual therapy and cervical proprioceptive retraining that the proprioceptive-mismatch route in Figure 8.1 prescribes.
The evidence gap for vestibular rehabilitation
Vestibular rehabilitation has a strong evidence base for peripheral vestibular disorders, but its evidence in cervicogenic dizziness specifically is much weaker. Lystad's review found insufficient data to recommend combining vestibular rehabilitation with manual therapy for cervicogenic dizziness, and noted that habituation drills designed for peripheral vestibular loss may be inappropriate when applied to a proprioceptive cause.42 The clinical implication: vestibular rehabilitation has its place in cervicogenic dizziness — predominantly in the chronic / centrally-reweighted patient (Route 4) — but it is not the first-line intervention for the typical proprioceptive presentation.
The full management pathway
Figure 8.2 — The management pathway. Phase 1 confirms diagnostic readiness (workup clean, differentials excluded). Phase 2 identifies which of the four mechanisms dominates this particular patient. Phase 3 routes to the treatment matched to that mechanism. Phase 4 reviews response at 6–12 weeks and escalates re-evaluation when progress is absent — at which point the most common discovery is that a different mechanism is actually in play, or that a co-existing diagnosis (most often vestibular migraine) was missed.
Four phases:
Pre-treatment — confirm that the diagnostic workup is complete (Module 5), differentials excluded (Module 7), and red flags addressed (Module 6). A cervicogenic diagnosis without this groundwork is not defensible.
Identify the dominant mechanism — using the Module 3 framework. History, examination, autonomic features, vascular features, and chronicity all contribute to this judgment.
Treatment route — chosen to match the dominant mechanism. The four routes are detailed in Figure 8.1; their colour-coding (teal=mechanical, red=RVAS, amber=central) carries through the pathway diagram.
Outcome assessment — at 6–12 weeks, with objective measures (Dizziness Handicap Inventory score, re-test of provocation tests, return to functional activities). Patients who are improving continue; those who are not return to Phase 2 — because most often, the actual answer is that a different mechanism is dominant or that a co-existing diagnosis was missed.
What to avoid — the four-route version
Iatrogenic harm in this space comes from applying the wrong treatment to the right diagnosis. The non-overlapping contraindications across the four routes deserve specific attention:
In RVAS: never apply high-velocity cervical manipulation. The rotational thrust is the very stimulus that provokes the artery. Gentle mobilisation appropriate to mechanism 1 is a different technique and a different patient.
In central reweighting / PPPD overlap: avoid pure manual therapy as a sole intervention. The mechanical work continues to have a role, but its centrality in this patient is over. Visual desensitisation and sensory-reweighting rehab carry the load.
In proprioceptive mismatch: avoid pure vestibular habituation drills. They treat a peripheral lesion the patient does not have, and may reinforce visual dependence — pushing the patient toward the chronic (Route 4) phenotype.
Across all routes: avoid long-term cervical collars and avoid stacked vestibular suppressants. Immobilisation reduces normal afferent flow; suppressants blunt central compensation. Both make the underlying problem worse rather than better.
Outcome measures worth tracking
For both individual patient care and any audit work, three measures form a minimum dataset:
Dizziness Handicap Inventory (DHI) — a 25-item patient-reported outcome covering functional, emotional, and physical impact. Sensitive to change over weeks to months.
Re-test of the provocation manoeuvres — at minimum, cervical torsion test reproducibility and SPNT/JPE (Module 4). If these have normalised, the proprioceptive mechanism is resolving.
Functional return-to-activity — work, sleep, driving, exercise. Patients understand and value these endpoints more than measurement-based ones.
Mechanism selection at the bedside
Mapping a patient to the dominant mechanism is the single most consequential judgment in this module. The clinical features that point to each route, with the highest-yield discriminator for each:
Route 1 (Proprioceptive mismatch) — patients with reproducible cervical findings: tender upper cervical segments, restricted segmental motion, positive cervical torsion test, abnormal SPNT or JPE. Discriminator: symptoms track the neck position in real time, not just position changes.
Route 2 (Vestibulo-sympathetic dominant) — patients whose autonomic features (nausea, palpitations, pallor, lightheadedness) are louder than the spinning sensation itself. Discriminator: a patient who describes feeling "ill" rather than "dizzy" during episodes.
Route 3 (RVAS) — patients with brainstem features (diplopia, dysarthria, drop attacks, transient hemisensory or hemimotor signs) reproducibly triggered by sustained head rotation. Discriminator: brainstem symptoms, not proprioceptive ones, during the provoked episode.
Route 4 (Central reweighting / chronic) — chronicity (≥3 months), visual dependence, exacerbation by busy visual environments and upright postures, prominent avoidance behaviour. Discriminator: symptoms now triggered by environments more than by neck movements.
When mechanisms co-exist
Most chronic patients carry contributions from more than one mechanism — typically a Route 1 baseline that has progressed into Route 4 features. The pragmatic clinical approach is to select the currently-dominant mechanism for the starting treatment plan, and to add a second route as an adjunct rather than starting with both at full intensity. Adding cervical sensorimotor training to a vestibular rehabilitation programme is reasonable; running both intensively in parallel without identifying the dominant driver is a known recipe for plateau and patient confusion.
When co-existing diagnoses are present
Co-existing diagnoses are the rule rather than the exception in patients who reach a tertiary service. The treatment sequence depends on which diagnosis is causing the most current disability:
Cervicogenic + vestibular migraine: treat the migraine first (prophylaxis, trigger control, triptans/gepants for attacks); reassess the cervicogenic features in 8–12 weeks. The frequent finding is that the cervicogenic features were partly migraine-driven, and the residual treatable cervical component is much smaller.
Cervicogenic + BPPV: reposition the BPPV first (Epley or Lempert manoeuvre); allow 1–2 weeks for settling; then start the Route 1 treatment with the position-evoked component removed.
Cervicogenic + PPPD: this is essentially Route 4. PPPD-specific treatment (vestibular rehab plus cognitive elements, SSRI/SNRI for some patients per Staab criteria38) takes precedence.
Cervicogenic + asymmetric audiovestibular features: stop and get the MRI brain with IAC protocol. Confirm schwannoma absence before continuing the cervicogenic treatment plan.
Pharmacology — what to use sparingly, what to avoid
Cervicogenic dizziness has no specific pharmacotherapy. The medications commonly prescribed deserve discrimination:
Vestibular suppressants (cinnarizine, prochlorperazine, meclizine, dimenhydrinate) — useful for brief symptomatic cover during severe flare-ups (24–72 hours). Beyond that they blunt central compensation, add anticholinergic load (particularly in older patients), and do nothing for the underlying mechanism. Limit duration actively.
Betahistine — has a role in Ménière's, not in cervicogenic dizziness. The trial evidence does not support an indication here.
Benzodiazepines — particularly bad in this population. Short-term anxiolysis at the cost of long-term dependency, blunted compensation, and fall risk in older patients. Reserve for the rare acute panic episode in the ER, not for outpatient management.
Anti-emetics — ondansetron or metoclopramide for autonomically-dominant presentations (Route 2). Cleaner side-effect profiles than first-generation antihistamines or scopolamine.
Migraine prophylaxis — when the workup or Module 7 engine suggests vestibular migraine is co-existing, a 2–3 month trial of a first-line prophylactic (propranolol, topiramate, amitriptyline, or for resistant cases a CGRP-targeted agent) is often diagnostically and therapeutically informative.
SSRI / SNRI — the Staab PPPD criteria list SSRIs/SNRIs as first-line pharmacotherapy for PPPD38, and they have a role in chronic Route 4 patients with substantial functional impairment.
NSAIDs and muscle relaxants — short-course for pain that is itself disturbing cervical spindle gain; not for chronic background management.
When to refer onward
The cervicogenic patient who does not improve at 6–12 weeks on a correctly-identified route is rarely a refractory cervicogenic patient — they are usually a different diagnosis in disguise. The escalation map:
To neurology — when central oculomotor signs have appeared, when episodes have evolved to look more like vestibular migraine, when red flags have developed, or when the patient is unresponsive to two routes adequately trialled.
To neuro-otology — when the workup needs advanced vestibular testing not previously performed (rotary-chair, subjective visual vertical, cVEMP/oVEMP), when a Ménière's spectrum is emerging, or when complex audiovestibular features require specialist interpretation.
To vascular neurology / neurosurgery — when RVAS is confirmed with structural correlate and surgical decompression is being considered.35
To vestibular physiotherapy — most patients, most of the time. The combined manual + sensorimotor programme that the evidence supports is rarely deliverable in the medical clinic alone.
To psychology / behavioural medicine — for chronic Route 4 patients with prominent avoidance, fear of falling, or co-existing anxiety/depression. The functional recovery in these patients is often more dependent on cognitive-behavioural work than on additional mechanical or vestibular intervention.
Patient expectation-setting
A consultation that ends without setting expectations is a consultation that has set them implicitly — and usually badly. Specific framings that recur as useful:
Realistic timelines — Route 1 patients should expect noticeable improvement at 4–6 weeks, with consolidation at 12 weeks. Route 4 patients should expect slower progress measured in months, with relapse risk during periods of stress, illness, or activity withdrawal.
Relapse-and-recovery framing — particularly in Route 4 patients, periods of regression are part of the trajectory, not failures of the treatment. Frame this explicitly before the first relapse occurs.
Active rather than passive language — workyou do, exercises you practice, positions you build tolerance to. The cervicogenic patient who sees themselves as the agent of their recovery improves better than the one who sees themselves as a recipient of interventions.
Honest acknowledgement of diagnostic uncertainty — particularly in light of the Bárány position statement, it is appropriate to be transparent that the mechanism is not fully proven and that diagnostic re-evaluation is part of the plan if response is poor. Patients respond well to this; it does not undermine confidence in care.
The clinical synthesis. Cervicogenic dizziness is a treatable condition for most patients who are correctly mechanism-mapped and correctly routed. The two recurring failure modes — missing a co-existing diagnosis and applying the wrong route to the right diagnosis — are largely preventable with the framework this module provides. The remaining art is in identifying the dominant mechanism for this specific patient, in this specific consultation, and revising that judgment honestly when progress fails to materialise.