What CDP is — and what it isn't

Reading levels: foundation · trainee · clinician · Anchor: introduction#what-is-cdp

Computerised Dynamic Posturography (CDP) is a functional assessment of balance. A patient stands on a force plate inside a movable visual surround while the apparatus systematically degrades each sensory input — vision, somatosensation, and vestibular cues — to reveal how the central nervous system weighs and integrates them.

Unlike calorics or the head impulse test, CDP does not localise a peripheral lesion. It answers a different question: given whatever organic substrate is present, how well can this patient stand? That makes it complementary to, not a replacement for, the rest of the vestibular work-up.

The test takes about 25 minutes in practised hands. It is non-invasive, well tolerated by most patients with a safety harness, and yields three protocols' worth of data — the Sensory Organisation Test (SOT), the Motor Control Test (MCT), and the Adaptation Test (ADT) — each interrogating a different layer of postural control.

Where CDP fits in the vestibular work-up

Reading levels: foundation · trainee · clinician · Anchor: introduction#where-it-fits

Think of the vestibular work-up as a stack. At the bottom sit anatomical tests — the head impulse test, calorics, VEMPs — which ask where is the lesion. Above them sit functional tests, of which CDP is the most informative single example. CDP asks how is the patient coping.

This functional layer matters because two patients with identical caloric weaknesses can have very different daily lives. One may have compensated fully and returned to work; the other may still be falling in the dark. CDP can distinguish them when the structural tests cannot.

CDP also plays a role in monitoring. Vestibular rehabilitation typically aims to improve sensory weighting and reduce visual dependence; serial CDP can document that improvement objectively in a way that's hard for symptom diaries alone.

The three protocols at a glance

Reading levels: foundation · trainee · clinician · Anchor: introduction#three-protocols

The Sensory Organisation Test presents six conditions that systematically remove or destabilise visual and somatosensory cues, forcing the patient onto residual inputs — and on conditions 5 and 6, onto vestibular input alone. It is the workhorse of CDP.

The Motor Control Test applies sudden platform translations to evoke automatic postural responses. Its key measurement is the latency of that response, which reflects long-loop brainstem-spinal pathways and is sensitive to central involvement.

The Adaptation Test uses repeated toes-up and toes-down platform rotations. The healthy nervous system rapidly damps the inappropriate sway response across trials — a cerebellum-dependent learning process. Failure of this adaptation is itself a useful finding.

Historical context: from Nashner to today

Reading levels: trainee · clinician · Anchor: introduction#history

The technique grew out of the work of Lewis Nashner at Good Samaritan Hospital in Portland through the late 1970s and early 1980s. Nashner's 1982 demonstration that postural sway could be selectively modulated by sway-referencing the surface and surround established the conceptual framework on which all modern CDP rests.

Commercial systems followed in the 1980s and 1990s, with NeuroCom (now Natus Balance Manager) dominating the field. The basic protocol — six SOT conditions, the MCT, and the ADT — has remained remarkably stable, although newer systems add force-plate-based stability tests and adaptive paradigms for rehabilitation use.

Through the 2000s and 2010s, large normative datasets (Ford-Smith 1995, Rine 2018 in paediatrics) refined the age-stratified cutoffs used today, and the Bárány Society consensus criteria for several vestibular disorders began to reference CDP findings as supportive evidence — though never as primary diagnostic criteria.

Limits of the technique

Reading levels: trainee · clinician · Anchor: introduction#limits

CDP is not a diagnostic test. A normal SOT does not exclude vestibular disease — BPPV in particular is essentially invisible between attacks. An abnormal SOT does not localise the lesion; the same pattern can arise from neuritis, schwannoma, ototoxicity, or a brainstem lacune.

The test is also susceptible to effort, anxiety, and expectation. The aphysiologic patterns described by Cevette and colleagues in 1995 reflect this — a CDP that looks too unusual to be biologically plausible is itself a finding, but a non-specific one.

Finally, CDP requires that the patient can stand for several minutes at a time. Severely deconditioned, frail, or acutely vertiginous patients may not tolerate the protocol. In those cases the rest of the bedside and laboratory work-up has to carry the load.

The three pillars: visual, vestibular, somatosensory

Reading levels: foundation · trainee · clinician · Anchor: anatomy#balance-system

Quiet standing depends on three sensory streams: vision, vestibular input from the inner ear, and somatosensation from receptors in the feet, ankles, joints, and muscles. The central nervous system integrates all three continuously, weighing each by how reliable it appears moment-to-moment.

When one input becomes unreliable — vision in the dark, somatosensation on a soft mattress, vestibular input after a unilateral lesion — the system re-weights toward the remaining streams. A healthy person can stand comfortably with any single input missing. Losing two simultaneously, or losing the only remaining functional input, is what produces falls.

CDP is built around exactly this re-weighting. By degrading each input in turn under controlled conditions, the test reveals which streams the patient actually relies on, and which are no longer carrying their share.

Vestibular labyrinth — relevant anatomy

Reading levels: foundation · trainee · clinician · Anchor: anatomy#vestibular-labyrinth

Each labyrinth contains three semicircular canals — anterior, posterior, and horizontal — and two otolith organs, the utricle and saccule. The canals sense angular head acceleration in three roughly orthogonal planes; the otoliths sense linear acceleration and head tilt relative to gravity.

For postural control on a fixed surface, the otoliths matter more than the canals. Subtle drifts of the head relative to gravity — what happens when the body sways — are the principal vestibular signal driving the corrective responses CDP measures. Canal input becomes more important during head turns and walking.

The afferents from both labyrinths converge on the vestibular nuclei in the dorsolateral pons and medulla, where they meet input from the cerebellum, the visual system, and the somatosensory pathways. The output goes to the spinal cord (vestibulospinal tracts), to the oculomotor nuclei (vestibulo-ocular reflex), and back up to cortex.

Central postural-control pathways

Reading levels: trainee · clinician · Anchor: anatomy#central-pathways

The descending postural pathways most relevant to CDP are the lateral vestibulospinal tract, originating from the lateral vestibular nucleus (Deiters' nucleus) and projecting bilaterally to anti-gravity muscles down the length of the spinal cord, and the reticulospinal tract, providing a more diffuse modulation of axial and proximal muscle tone.

Cerebellar input modulates both. The flocculus and paraflocculus tune the vestibulo-ocular reflex; the anterior vermis and fastigial nucleus tune postural responses; the posterior vermis handles trunk control. Cerebellar lesions in any of these regions can produce abnormal CDP findings, though the patterns differ in detail.

Cortical input is more sparing but real. Frontal cortex contributes to anticipatory postural adjustments; parietal cortex to the integration of vestibular and somatosensory cues. Cortical lesions rarely produce isolated CDP abnormalities, but they can shape the pattern in patients with multifocal disease.

Long-loop vs short-loop reflexes

Reading levels: clinician · Anchor: anatomy#long-loop

Short-loop reflexes — the monosynaptic stretch reflex and its polysynaptic siblings — operate at roughly 30–50 ms latency and are mediated entirely within the spinal cord. They are too fast and too local to account for the responses CDP measures.

Long-loop reflexes traverse brainstem and possibly cerebellar circuitry, and emerge at 100–200 ms after a perturbation. The MCT response, at 120–150 ms in healthy adults, is squarely in this range. The latency reflects the round-trip transit time through the long loop, not the strength of the response.

Prolongation of MCT latency therefore implicates the long loop specifically — meaning the brainstem, spinal cord, or the major descending tracts. Peripheral neuropathy can sometimes prolong latency by slowing the afferent or efferent limb, but more often it reduces amplitude (the response is weak rather than late). A latency over 170 ms in a patient with intact peripheral nerves is a substantial flag for central involvement.

Sensory weighting and re-weighting

Reading levels: foundation · trainee · clinician · Anchor: sensory-integration#weighting

Sensory weighting is the term for how much the brain trusts each balance input. In a healthy person standing on solid ground in good light, vision and somatosensation are weighted heavily because they're cheap and reliable; vestibular input is weighted more lightly, held in reserve.

When the environment changes — the lights go out, the ground softens — the brain re-weights within seconds. Vision gets discounted in the dark; somatosensation gets discounted on a foam pad; vestibular input picks up the slack. Healthy adults do this re-weighting unconsciously and continuously.

When the re-weighting fails — either because the alternative input is damaged (vestibular loss) or because the central re-weighting machinery is broken (some central disorders, or PPPD's hyper-reliance on vision) — the patient cannot smoothly adapt to changing sensory environments. They feel unsteady in dim corridors, busy supermarkets, or on uneven ground.

Sensory conflict resolution

Reading levels: foundation · trainee · clinician · Anchor: sensory-integration#conflict

When two sensory streams disagree — say, vision tells the brain the body is upright while vestibular input says it's tilting — the brain must resolve the conflict. Healthy resolution involves down-weighting the stream judged less reliable in the current context.

On condition 3 of the SOT, the surround sways with the patient. Vision now reports that the body is stationary (because the visual world moves in lock-step), while vestibular and somatosensory input correctly report movement. A healthy subject down-weights vision and stays upright. A subject with visual preference — hyper-reliance on vision — cannot down-weight, and sways in line with the moving surround.

This is the physiological basis of the PREF ratio derived from the SOT. Elevated PREF identifies patients whose central re-weighting is biased toward vision, a pattern characteristic of PPPD and of some recovering peripheral vestibulopathies.

Ankle and hip postural strategies

Reading levels: foundation · trainee · clinician · Anchor: sensory-integration#ankle-hip

Two postural strategies handle perturbations. The ankle strategy uses ankle torque to keep the centre of mass over the base of support; the body rotates about the ankle joints, with the rest of the body kept relatively straight. It works well for small, slow perturbations on a stable surface.

The hip strategy recruits hip torque, bending the body at the waist. The trunk and arms move while the legs stay relatively still. It is recruited for larger or faster perturbations, or when the support surface is too narrow or too soft for ankle torque to suffice.

Healthy adults transition smoothly between the two. Loss of the ankle strategy — for example after lower-leg amputation or severe peripheral neuropathy — forces reliance on hip strategy and shows up as elevated sway on CDP. Loss of the hip strategy is rarer but can occur with hip pathology or with central lesions affecting trunk control.

Limits of stability

Reading levels: trainee · clinician · Anchor: sensory-integration#los-anatomy

The limit of stability (LOS) is the boundary of the cone of sway within which the body can remain upright without taking a step. For a typical adult on a stable surface, this is about 12.5 degrees anteriorly, 6.5 degrees posteriorly, and 8 degrees laterally — though the anterior–posterior asymmetry is what CDP scoring uses.

Within the LOS, the patient can correct by ankle or hip torque alone. Exceed it and a step (or a fall) becomes inevitable. CDP scoring expresses sway as a fraction of the LOS, with an equilibrium score of 100 meaning negligible sway and 0 meaning the cone was breached.

The 12.5 degree figure is an idealised anterior limit. Real-world LOS depends on the patient's height, stance width, footwear, and lower-limb strength. Most commercial systems use the 12.5 degree default, accepting some imprecision in exchange for cross-patient comparability.

Predictive vs reactive postural control

Reading levels: clinician · Anchor: sensory-integration#predictive

Postural control has two distinct modes. Reactive control responds to a perturbation that has already happened — the MCT measures it. Predictive control anticipates a perturbation that is about to happen, and pre-tunes muscle tone and posture accordingly.

Predictive control depends on cortical, cerebellar, and basal-ganglia circuits that learn from prior experience. Parkinson's disease classically impairs predictive control: patients can react to a push, but they don't prepare for an expected one, producing the characteristic backward fall when expecting a pull.

Standard CDP does not isolate predictive control; the test paradigm is largely reactive. Adaptive paradigms in research use repeated perturbations with predictable cues to probe predictive control specifically, but these are not yet part of routine clinical protocols.

The force plate

Reading levels: foundation · trainee · clinician · Anchor: technique#force-plate

The force plate is a rigid platform suspended on four strain-gauge transducers, one at each corner. From the four vertical forces, the system computes the centre of pressure — the single point at which the patient's vertical ground-reaction force can be considered to act.

As the patient sways, the centre of pressure shifts beneath them. Anterior-posterior shifts dominate quiet standing and are the principal CDP signal. Mediolateral sway is smaller in most conditions but matters in some disorders (cerebellar truncal ataxia in particular).

Commercial systems sample the force plate at 100 Hz or higher. The signal is filtered to remove cardiac and respiratory artefact and converted into degrees of body sway by assuming a single-segment inverted-pendulum model — a reasonable approximation for quiet stance though not for postural strategies that bend at the waist.

Sway-referencing of surface and surround

Reading levels: foundation · trainee · clinician · Anchor: technique#sway-referencing

Sway-referencing is the central technical innovation that makes CDP possible. The platform (or visual surround) rotates in real time about the ankle axis to track the patient's body sway, holding the support surface — from the patient's perspective — in a constant relationship to the body.

The patient cannot feel the rotation directly. What they experience is that the floor no longer provides reliable information about which way is down; if they lean forward, the platform rotates forward to match them, so the ankle joint angle doesn't change. Somatosensation now reports stationarity even as the body falls.

The same trick can be applied to the visual surround — rotating the walls in lock-step with body sway so that the visual flow field reports no movement. This is what produces conditions 3 and 6 of the SOT, where vision becomes uninformative.

Setup, calibration, and safety harness

Reading levels: foundation · trainee · clinician · Anchor: technique#calibration

Setup takes about five minutes. The patient stands on the force plate in stocking feet, with foot position marked using a height-based template — taller patients stand with their feet further apart, so that the centre of mass falls in a consistent location relative to the force-plate origin.

A safety harness, attached to an overhead frame, prevents falls without supporting the patient's weight during the trial. A correctly fitted harness is taut enough to catch but loose enough that the patient does not feel suspended; some systems include a tension sensor to flag trials in which the harness took load.

Calibration confirms that the force plate's centre is correctly registered and that the surround motors respond linearly. Daily zeroing is recommended; full calibration every few months or after any service event.

Common artefacts and pitfalls

Reading levels: trainee · clinician · Anchor: technique#artefacts

Harness assist is the commonest artefact: a patient leans into the harness, transferring some weight to the overhead frame. Equilibrium scores rise spuriously. Modern systems flag this via tension sensors; older systems require operator vigilance and re-instruction.

Step responses end the trial early. Some scoring conventions credit a step as a fall (equilibrium score 0); others split the response between completed sway and the step itself. Know which convention your system uses before comparing scores across centres.

Foot placement drift across the six SOT conditions changes the effective base of support and can shift the equilibrium scores by several points. Re-mark foot position between conditions if the patient steps off and back on.

Anxiety stiffening reduces sway by recruiting co-contraction of agonist and antagonist muscles. The resulting scores are high but the strategy is non-physiological and shows up in strategy analysis (where it survives) as flat ankle traces with abnormally low sway frequencies.

Normative databases and device differences

Reading levels: trainee · clinician · Anchor: technique#norms

Normative data are device-specific. NeuroCom/Natus, Bertec, and other manufacturers each publish their own age-stratified norms; cross-device comparison of raw scores is hazardous. Always interpret a patient's data against the norms supplied with the device that recorded them.

Age effects are real and substantial. Composite scores drop by roughly 0.4 points per year after age 60, with most of the drop on conditions 5 and 6. A composite of 65 in a 75-year-old may be within normal limits; the same score in a 30-year-old is clearly abnormal.

Paediatric norms are sparser but have improved. Rine and colleagues (2018) published age-stratified data for children 5–18, showing adult-like patterns by about age 7 with gradual maturation of conditions 5 and 6 through adolescence.

Overview of the SOT

Reading levels: foundation · trainee · clinician · Anchor: sot#overview

The Sensory Organisation Test is the most-used CDP protocol. It consists of six conditions, each repeated for three 20-second trials, for a total of eighteen trials and about 12 minutes of recording time including rest periods.

Across the six conditions, vision is either available (eyes open) or removed (eyes closed), and the support surface and visual surround are either fixed or sway-referenced. The combinations systematically isolate each sensory stream.

The output of each trial is an equilibrium score from 0 to 100, with 100 meaning negligible sway and 0 meaning the patient's sway exceeded the limit of stability — operationally, a fall. The three trial scores per condition, the six condition means, and the overall composite all feed into pattern interpretation.

The six conditions

Reading levels: foundation · trainee · clinician · Anchor: sot#conditions

Condition 1: eyes open, fixed surface, fixed surround. The baseline. All three sensory inputs are reliable; healthy adults score above 90.

Condition 2: eyes closed, fixed surface. Vision removed; somatosensation and vestibular input remain reliable. Most healthy adults still score above 85.

Condition 3: eyes open, sway-referenced surround. The visual surround tracks body sway, making vision misleading. Somatosensation and vestibular input are unaffected. Healthy adults score in the mid-80s.

Condition 4: eyes open, sway-referenced surface. The floor tracks body sway, making somatosensation misleading. Vision and vestibular input are reliable. Scores drop into the mid-70s in healthy adults.

Condition 5: eyes closed, sway-referenced surface. Vision absent, somatosensation misleading. Only vestibular input remains. This is the hardest pure-input condition and the most sensitive to peripheral vestibular loss; healthy adults score in the 60–70 range.

Condition 6: eyes open, sway-referenced surface and surround. Both vision and somatosensation are misleading. Only vestibular input remains, but the patient must also ignore the false visual signal — harder than condition 5 cognitively, though the pure sensory information is identical.

Equilibrium score calculation

Reading levels: trainee · clinician · Anchor: sot#equilibrium-score

The equilibrium score is calculated from the peak anterior-posterior sway during the 20-second trial, expressed relative to the 12.5 degree limit of stability. The standard formula is EQS = (1 − θ_peak / 12.5) × 100, clamped to the [0, 100] range.

If the patient reaches or exceeds the limit of stability — or steps, or relies on the harness — the trial is scored as a fall (EQS = 0). Some clinicians report fall trials separately rather than averaging them into the condition mean, as a single fall in three trials carries different information than three near-falls.

A score of 100 is rarely seen in practice. Even still, healthy subjects sway by 1–2 degrees during quiet stance; 100 would require zero sway, which is non-physiological. Scores in the high 90s are typical for healthy condition 1 and 2 trials.

Composite score and age norms

Reading levels: foundation · trainee · clinician · Anchor: sot#composite

The composite score is a weighted average across all eighteen trials. It is the single number most often quoted in clinical reports, and it correlates well with global functional balance.

Age-matched lower limits of normal are roughly 70 in young and middle-aged adults, dropping to about 60 by age 75. The exact cut-off depends on the device's normative database — use the one supplied with the system that recorded the test.

A reduced composite is sensitive but non-specific. It tells you the patient has impaired balance, but not why. The condition-level pattern and the sensory ratios carry the diagnostic information; the composite is the summary metric.

Sensory analysis ratios

Reading levels: trainee · clinician · Anchor: sot#ratios

Four sensory ratios are derived from the SOT conditions, each isolating one sensory stream:

Somatosensory (SOM) = C2 / C1. How well the patient stands without vision, normalised to baseline. Normal values exceed 0.90.

Visual (VIS) = C4 / C1. How well the patient uses vision when somatosensation is unreliable. Normal values exceed 0.75.

Vestibular (VEST) = C5 / C1. Pure vestibular reliance. Normal values exceed 0.60. This is the ratio most often abnormal in peripheral vestibular disease.

Visual preference (PREF) = (C3 + C6) / (C2 + C5). Detects hyper-reliance on vision: a patient who falls when vision becomes misleading even though somatosensation/vestibular input are intact. Normal values are below 0.92. Elevated PREF is supportive of PPPD.

Pattern interpretation

Reading levels: foundation · trainee · clinician · Anchor: sot#patterns

Five canonical patterns emerge from the SOT, named for which conditions are abnormal:

Vestibular pattern: selective reduction on C5 and C6, all other conditions normal. Classical signature of acute peripheral vestibular loss — neuritis, an uncompensated schwannoma, or the acute phase of Ménière's. The most common pattern in vestibular clinic populations.

Visual-vestibular pattern: reduction on C4, C5, and C6 — the conditions in which somatosensation is unreliable. Park and colleagues (2017) found this in about 24% of acute neuritis cases.

Surface-dependent pattern: reduction on conditions with sway-referenced surfaces (C4–C6) but spared performance with reliable somatosensation. Seen in peripheral neuropathy where vestibular function may be normal but somatosensation is unreliable.

Visual preference pattern: PREF ratio elevated, with or without reductions on C3 and C6. Characteristic supportive feature of PPPD.

Aphysiologic pattern: easier conditions paradoxically worse than harder ones (the Cevette inversion), or extreme inter-trial variability, or patterns that don't fit any physiologically-plausible combination. Non-specific; warrants clinical correlation rather than diagnosis.

Strategy analysis (ankle vs hip)

Reading levels: clinician · Anchor: sot#strategy-analysis

Strategy analysis reports the proportion of corrective effort taken up by ankle versus hip torque during each trial. The system infers this from the shape of the sway trace — slow, smooth sway is ankle strategy; fast, abrupt sway is hip strategy.

Pure ankle strategy approaches 100 in conditions 1 and 2 in healthy adults. As conditions become more challenging, hip strategy is recruited and the score drops. In condition 5 in young healthy adults, strategy scores around 60–70 are typical.

Three abnormal patterns emerge. Premature hip recruitment — low strategy scores even in conditions 1 and 2 — suggests reduced confidence in ankle torque, often from lower-limb pathology or fear of falling. Failure to recruit hip strategy — strategy scores stuck near 100 even in difficult conditions — suggests trunk-control deficits, sometimes seen in cerebellar disease. Co-contraction stiffening, an anxiety-driven pattern, produces high strategy scores with non-physiological sway frequencies and should be flagged separately.

Overview of the MCT

Reading levels: foundation · trainee · clinician · Anchor: mct#overview

The Motor Control Test measures the automatic postural response to a sudden horizontal translation of the platform. It is the protocol most sensitive to central involvement, because the response transit time — the latency — reflects long-loop brainstem-spinal pathways.

The test takes about three minutes. The patient stands quietly while the platform delivers six perturbations: small, medium, and large translations in both forward and backward directions, with three trials at each amplitude-direction combination.

Three measurements come out of each trial: latency (in milliseconds), amplitude (in some systems expressed as response strength relative to platform displacement), and weight symmetry (the relative load on each leg during the response).

Translations: amplitudes and directions

Reading levels: foundation · trainee · clinician · Anchor: mct#perturbations

Translation amplitudes are scaled to patient height. A small translation moves the platform about 1.5 cm in a young adult; large translations are about 5.5 cm. The translation completes in roughly 400 ms, with peak velocity in the middle of that window.

When the platform translates backward, the patient's centre of mass lags behind — they begin to fall forward. The corrective response is a backward sway driven by anterior calf muscles. Forward platform translations produce the opposite response.

The amplitude scaling lets the test distinguish weak responses (small amplitude on a large perturbation) from absent responses (no response at any amplitude). It also reveals exaggerated responses — over-scaling — which suggests anxious hypersensitivity rather than organic disease.

Latency

Reading levels: foundation · trainee · clinician · Anchor: mct#latency

Latency is measured from platform-translation onset to the first deflection of the patient's centre of pressure in the corrective direction. Healthy adults respond at 120–150 ms; the lower limit of normal is around 110 ms and the upper limit around 165 ms.

Latencies above 170 ms in a patient with intact peripheral nerves implicate the long-loop pathway. The most informative single lesion is a brainstem stroke involving the descending vestibulospinal or reticulospinal tracts; less dramatically, cerebellar disease and demyelination can also prolong latency.

Peripheral neuropathy can prolong latency by slowing the afferent or efferent limb, but more often reduces amplitude. A combination of prolonged latency and reduced amplitude with confirmed neuropathy is consistent with the peripheral diagnosis. The same combination in a patient with intact peripheral nerves is a flag for central pathology.

Amplitude scaling

Reading levels: foundation · trainee · clinician · Anchor: mct#amplitude-scaling

Amplitude scaling describes how the response size varies with perturbation size. A healthy adult produces a small response to a small perturbation and a large response to a large one, scaled approximately linearly.

Under-scaling — responses too small for the perturbation — occurs in peripheral neuropathy, in late-stage central disorders, and in some patients with chronic deconditioning. Functionally, it represents an inadequate corrective force.

Over-scaling — responses too large for the perturbation — is more characteristic of anxious patients, including those with PPPD. The brain over-corrects, sometimes producing oscillation between the initial response and a counter-response. Over-scaling is one of the supportive CDP features of PPPD listed in the Bárány Society criteria.

Weight symmetry

Reading levels: trainee · clinician · Anchor: mct#weight-symmetry

Weight symmetry compares the load on the two legs during the corrective response. The convention varies between systems; one common form expresses asymmetry as (R − L) / (R + L), with healthy values near zero (perfect symmetry).

Asymmetric responses can reflect lower-limb pathology — pain, weakness, recent surgery — or hemiparesis. They are less specific than latency or amplitude findings and need clinical correlation.

Persistent asymmetry after a unilateral vestibular lesion is unusual on the MCT, because the perturbation drives a bilateral postural response that doesn't depend on labyrinthine input. If asymmetry is present in this setting, look for an additional lower-limb cause.

Pattern interpretation

Reading levels: trainee · clinician · Anchor: mct#interpretation

Three latency-based patterns matter most. Normal latency, normal amplitude: the long-loop pathway is intact; central pathology is unlikely. Prolonged latency, any amplitude: long-loop involvement; central or peripheral nerve cause; correlate clinically. Normal latency, reduced amplitude: peripheral or musculoskeletal cause more likely.

Amplitude scaling adds nuance. Linear under-scaling across amplitudes suggests neuropathy or deconditioning; non-linear over-scaling at small amplitudes with normal scaling at large amplitudes is a hint toward anxious hypersensitivity.

The MCT alone rarely makes a diagnosis. Its value lies in confirming or refuting central involvement when the SOT shows a peripheral pattern, and in objectively documenting the hypersensitivity profile that may underlie PPPD.

Overview of the ADT

Reading levels: foundation · trainee · clinician · Anchor: adt#overview

The Adaptation Test measures the central nervous system's ability to learn from a perturbation and damp the response across trials. It is the CDP protocol most sensitive to cerebellar involvement.

The test consists of two series of five trials each: toes-up rotations of the platform, and toes-down rotations. Each rotation lifts or drops the front of the platform by about 8 degrees over 400 ms.

The response to a toes-up rotation is initially exaggerated — the patient sways backward as if pushed. By the fifth trial, the healthy central nervous system has learned that the perturbation is harmless and damps the response. Sway energy on trial five is typically 40–60% of trial one.

Toes-up and toes-down rotations

Reading levels: foundation · trainee · clinician · Anchor: adt#toes-up-down

Toes-up rotation: the front of the platform rises. The patient's body initially leans forward (relative to the new platform tilt), then sways backward as a corrective response. Healthy adults overshoot slightly on trial one, then progressively adapt.

Toes-down rotation: the front of the platform drops. The patient leans backward, then sways forward correctively. The adaptation pattern is similar to toes-up, although mean sway energies are typically a touch lower.

Some clinicians report toes-up and toes-down separately because asymmetric adaptation can occur — typically with toes-down adapting faster than toes-up. The interpretation of mild asymmetry is not standardised; treat it as a soft finding unless one direction is clearly non-adapting.

Sway energy

Reading levels: trainee · clinician · Anchor: adt#sway-energy

Sway energy is the integrated sway across the trial, weighted by frequency. It captures both the amplitude and the timing of the postural response in a single number. Higher values mean more total sway; the system normalises to a baseline so that direct trial-to-trial comparison is meaningful.

On a healthy first trial, sway energy is high — the patient hasn't yet learned the perturbation. By trial five, sway energy has typically dropped substantially: a fall of 40–60% from trial one is normal.

A flat trajectory across all five trials — sway energy on trial five roughly equal to trial one — is the abnormal finding. It means the central adaptation machinery is not learning. Cerebellar disease is the most specific cause; some central white-matter disorders can produce similar patterns.

Central adaptation

Reading levels: foundation · trainee · clinician · Anchor: adt#adaptation

Adaptation requires cerebellar circuitry, particularly the anterior vermis and the interposed nuclei. These regions receive input about the perturbation (from somatosensory afferents and the vestibular system) and tune the descending motor command on subsequent trials.

The time-course of adaptation is rapid: most of the gain occurs between trials one and three. A patient who reaches their floor sway energy by trial three and stays flat for trials four and five is adapting normally; a patient whose trial-five sway is essentially identical to trial one is not.

Adaptation is reversible in the short term — fully-adapted patients tested again after a half-hour rest will show a partial return of trial-one responses, though usually not the full pre-adaptation level. Day-to-day learning probably accumulates more slowly via the same circuitry.

Pattern interpretation

Reading levels: trainee · clinician · Anchor: adt#interpretation

Three patterns matter. Normal adaptation: clear downward trend across the five trials in both directions; trial five well below trial one. Most patients fall here.

Failure of adaptation: flat or even upward trend. Strongly suggestive of cerebellar disease, especially if accompanied by clinical signs (dysmetria, ataxia, gaze-evoked nystagmus). Demyelinating disease, late-stage Parkinson's, and some metabolic encephalopathies can also produce non-adapting ADT.

Direction asymmetry: adaptation in one direction but not the other. The interpretation here is uncertain; mild asymmetry probably reflects normal inter-trial variability, but a striking asymmetry should be flagged for follow-up.

What the Limits of Stability test measures

Reading levels: foundation · trainee · clinician · Anchor: limits-of-stability#what

The Limits of Stability (LoS) test measures how far a person can deliberately lean — shifting their centre of gravity towards a target — without stepping, reaching, or falling. Where the SOT asks how well a patient stays still under sensory challenge, the LoS asks how confidently they can move to the edge of their base of support and back.

The patient stands on the force plate and watches a screen showing their real-time centre-of-gravity cursor and a set of targets arranged around them, typically at eight compass points. On a cue, they lean to move the cursor onto each target as quickly and accurately as they can, then hold.

Because it probes volitional postural control rather than reactive or sensory-organisation control, the LoS complements the SOT, MCT and ADT. It is especially informative in cautious, deconditioned, and fear-of-falling patients whose quiet-stance scores look deceptively normal.

The five LoS parameters

Reading levels: foundation · trainee · clinician · Anchor: limits-of-stability#parameters

Reaction time — the delay between the go cue and the first movement of the centre of gravity. Prolonged in central disorders, Parkinsonism, and with cautious set.

Movement velocity — the average speed of the intentional lean towards the target. Slow movement velocity is one of the most sensitive markers of reduced postural confidence.

Endpoint excursion — how far the centre of gravity travels on the first attempt toward the target, expressed as a percentage of the theoretical maximum. Maximal excursion is the farthest reached during the whole trial.

Directional control — how much of the movement is on-axis toward the target versus wasted off-axis sway. Low directional control suggests poor coordination or a tremor-laden trajectory.

Reading the LoS — constriction patterns

Reading levels: trainee · clinician · Anchor: limits-of-stability#interpretation

The hallmark abnormality is a constricted stability envelope — endpoint and maximal excursions fall short of the target ring in some or all directions, so the leaning area the patient is willing to use is smaller than normal.

Direction matters. A symmetric global constriction is typical of generalised cautious set, deconditioning, and fear of falling. A directionally asymmetric constriction — short excursions only toward one side — points to a lateralised deficit, hemiparesis, or a painful or unstable limb.

Reading the temporal parameters alongside the spatial ones separates mechanisms: slow reaction time with preserved excursion suggests a central initiation problem, whereas normal reaction time with poor directional control suggests an execution or coordination problem.

Fall risk and rehabilitation targets

Reading levels: trainee · clinician · Anchor: limits-of-stability#clinical

LoS metrics correlate with fall risk in older adults and in neurological populations, and a constricted, slow envelope is a practical, objective marker of the fear-of-falling cycle that perpetuates disability after a first fall.

Because the parameters are intuitive and visual, the LoS doubles as a rehabilitation tool: patients can see their cursor reach further and faster over a course of therapy, which is motivating and provides an objective progress record that pairs well with serial SOT composites.

The test is not diagnostic of any single disease. Its value is functional — it quantifies the volitional-control component of balance that quiet-stance testing misses, and it gives rehabilitation a concrete, trainable target.

What 'normal' looks like

Reading levels: foundation · trainee · clinician · Anchor: normal#overview

The healthy adult SOT shows a stereotyped pattern: scores in the high nineties on conditions 1 and 2, sliding gradually into the high seventies for condition 4, and bottoming out in the low-to-mid sixties for conditions 5 and 6.

The composite score sits between 70 and 85 in healthy adults under 60. Sensory ratios sit at: SOM ≥ 0.90, VIS ≥ 0.75, VEST ≥ 0.60, PREF ≤ 0.92. None of these cut-offs is exact — they depend on the normative database used.

MCT latencies fall in the 120–150 ms range with symmetric responses scaling appropriately to perturbation amplitude. The ADT shows clear adaptation, with trial five sway energy 40–60% lower than trial one in both toes-up and toes-down series.

Age effects on CDP measures

Reading levels: foundation · trainee · clinician · Anchor: normal#age-effects

Age affects every CDP measure, but unevenly. The composite score drops by roughly 0.4 points per year after age 60, with most of the decline concentrated on conditions 5 and 6.

MCT latencies prolong modestly with age — about 1–2 ms per decade after 50 — and amplitude scaling becomes slightly less linear. The principal age-related MCT change is increased trial-to-trial variability rather than a shift in mean latency.

ADT adaptation appears to be relatively preserved with healthy ageing. Older adults adapt more slowly across the five trials but still show clear adaptation by trial five. Failure of adaptation in an older patient is not a normal age effect and should be investigated.

Test-retest reliability and learning effects

Reading levels: trainee · clinician · Anchor: normal#test-retest

Test-retest reliability of the composite score is good — intraclass correlations of 0.7–0.8 over a few weeks in healthy adults — but individual condition scores are less reliable, particularly C5 and C6. Composite is the most stable metric for tracking patients over time.

Learning effects are real but small. Repeat testing within hours can improve scores by a few points, particularly on conditions 5 and 6. By 24 hours, the learning effect has largely washed out. Schedule serial CDPs at least a week apart to minimise this confound.

Diurnal variation is unhelpful but real — late-day tests typically show slightly lower scores than morning tests, attributed to fatigue. The effect is small (a point or two on the composite) and rarely clinically significant.

Bilateral and unilateral loss — overview

Reading levels: foundation · trainee · clinician · Anchor: vestibular-loss#overview

Vestibular loss can be unilateral or bilateral, partial or complete, acute or chronic. The CDP pattern depends on all four axes. Acute losses produce dramatic patterns; chronic compensated losses can look near-normal. Bilateral losses produce more severe and stereotyped patterns than unilateral losses.

CDP is sensitive to severity but not to side. A unilateral loss does not produce a directionally-asymmetric SOT — body sway is roughly symmetric in either direction on every condition, because postural control draws on bilateral vestibular input integrated centrally.

What CDP does well in this group: confirms that a structural lesion has functional consequences, monitors compensation over time, and identifies the rare patient whose CDP pattern is much worse than their imaging would suggest (raising the question of additional central involvement).

Unilateral peripheral loss

Reading levels: foundation · trainee · clinician · Anchor: vestibular-loss#unilateral

Acute unilateral vestibular loss — neuritis, labyrinthitis, the early hours after labyrinthectomy — produces the classical vestibular pattern: selective reduction on C5 and C6, falls common on C5, with C1–C4 typically intact. About 70% of acute neuritis patients show this pattern (Park 2017).

Chronic compensated unilateral loss is much milder. Most patients with caloric weaknesses of 30–50% have entirely normal CDPs once central compensation has occurred, usually within three to six months. A patient with documented unilateral weakness but persistently abnormal CDP at six months has either failed to compensate, has an additional problem, or has developed a secondary disorder like PPPD.

The asymmetry of caloric findings does not need to predict the asymmetry of CDP findings, because the CDP test does not isolate one side. A 70% unilateral weakness can produce a normal CDP if compensation is good, and a 20% weakness can produce an abnormal CDP if compensation has failed.

Bilateral vestibulopathy

Reading levels: trainee · clinician · Anchor: vestibular-loss#bilateral

Bilateral vestibulopathy produces the most severe and stereotyped CDP pattern in the disorder library. C5 and C6 are essentially zero — the patient cannot stand on a sway-referenced surface without vestibular input. C4 is often reduced too; even C2 can be borderline in severe cases.

The Bárány Society 2017 criteria for bilateral vestibulopathy require bedside HIT abnormality bilaterally plus reduced caloric responses bilaterally (sum of maximum slow-phase velocities < 6°/s per side). CDP supports the diagnosis but is not part of the criteria.

Aminoglycoside ototoxicity is the most common identifiable cause. Other causes include autoimmune inner-ear disease, recurrent Ménière's affecting both labyrinths, and idiopathic progressive vestibulopathy. CDP can document the severity baseline and track rehabilitation response.

Central compensation and recovery

Reading levels: trainee · clinician · Anchor: vestibular-loss#compensation

Central compensation after unilateral vestibular loss occurs through several mechanisms: rebalancing of resting vestibular tone between the two vestibular nuclei, re-weighting of sensory inputs toward intact vision and somatosensation, and learning of new motor patterns through cerebellar plasticity.

The time course is variable. Most patients show substantial CDP improvement within four to six weeks of an acute lesion, and reach a plateau by three to six months. A patient still showing a severe vestibular pattern at six months is unusual and warrants a careful look at compliance with vestibular rehabilitation, central involvement, and PPPD overlay.

Vestibular rehabilitation appears to accelerate compensation. Serial CDP can document this objectively — a patient whose composite rises from 50 to 70 over eight weeks of therapy has measurably improved, regardless of whether their symptoms have resolved completely.

Archetypal signature

The textbook 'vestibular pattern' — selective failure on C5 and C6 with preserved performance on C1–C4. Reflects an inability to use vestibular input when both vision and somatosensation are sway-referenced or unavailable.

Signature jitter: 5 · MCT latency: 145 ms · ADT toes-up: [70, 58, 48, 42, 38]

Acute vestibular neuritis

Reading levels: foundation · trainee · clinician · Anchor: vestibular-neuritis#overview

Vestibular neuritis is an acute peripheral vestibulopathy presenting with sudden severe vertigo lasting days to weeks, in a patient without auditory symptoms and without central neurological signs. It is most often attributed to inflammation of the vestibular nerve, possibly viral.

The acute findings are unilateral: a positive head impulse test to one side, caloric weakness on the same side, contralesional spontaneous nystagmus that is suppressed by visual fixation, and no hearing loss. The HINTS exam (head impulse, nystagmus, test of skew) reliably distinguishes neuritis from a posterior-circulation stroke in the acute setting.

Recovery is the rule, although a minority of patients develop PPPD or chronic dizziness even after the peripheral lesion has compensated. CDP can document the acute lesion and track its compensation.

Typical CDP pattern

Reading levels: foundation · trainee · clinician · Anchor: vestibular-neuritis#cdp-pattern

Park and colleagues (2017) examined CDP in 87 patients with acute vestibular neuritis, all tested within 7 days of symptom onset. Seventy percent had abnormality on conditions 5 or 6, or both. Falls on condition 5 were common.

The most frequent SOT pattern was the classical vestibular pattern: selective reduction on C5 and C6 with C1–C4 intact (about 46% of patients). The second most frequent was the visual-vestibular pattern, with reduction extending to C4 (about 24% of patients).

MCT findings in acute neuritis are typically normal in latency and amplitude. The ADT is usually intact — adaptation is a central process and the lesion is peripheral. Findings outside this pattern, particularly prolonged MCT latency or non-adapting ADT, should prompt a search for central involvement.

Evolution of CDP findings over time

Reading levels: trainee · clinician · Anchor: vestibular-neuritis#evolution

Over the first six weeks, the CDP pattern attenuates. The classical vestibular pattern softens as central compensation rebalances vestibular tone; condition 5 scores rise from near-zero into the 30s and 40s, then into the 50s and 60s by three months.

By six months, most uncomplicated neuritis patients have essentially normal CDPs even when caloric weakness persists. This dissociation between structural and functional recovery is the key reason CDP and calorics report complementary, not redundant, information.

Patients who do not show this trajectory — those still with a severe vestibular pattern at three months — warrant a careful re-look. Co-existing PPPD, fear-avoidance behaviour, central involvement, and incomplete rehabilitation are the principal possibilities.

Archetypal signature

In acute VN, Park et al. (2017) found that ~70% of patients show abnormality on C5 and/or C6. The pattern is more severe than chronic compensated loss, with frequent falls on C5/C6 and high intertrial variability reflecting the uncompensated state.

Signature jitter: 7 · MCT latency: 142 ms · ADT toes-up: [78, 70, 62, 56, 52]

Ménière's disease overview

Reading levels: foundation · trainee · clinician · Anchor: menieres#overview

Ménière's disease is defined clinically (Bárány Society 2015) by recurrent vertigo attacks lasting 20 minutes to 12 hours, with audiometrically documented low-to-medium frequency sensorineural hearing loss on the affected ear and fluctuating aural symptoms (fullness, tinnitus) on the same side.

The disease evolves over years. Early on, attacks are infrequent and hearing recovers between them. Over time, attack frequency may increase, hearing loss becomes permanent, and the vestibular function on the affected side gradually declines toward stable hypofunction.

CDP plays a supporting role. It is not part of the diagnostic criteria; the clinical picture and audiometric documentation carry the diagnosis. CDP can document interictal vestibular function, track fluctuation across attacks, and identify the rare patient whose Ménière's-like picture is actually something else.

Interictal CDP

Reading levels: foundation · trainee · clinician · Anchor: menieres#interictal

Between attacks, CDP in Ménière's typically shows a mild-to-moderate vestibular pattern: condition 5 and 6 scores below age-matched norms but not zero, with C1–C4 generally intact. The pattern is similar in shape to vestibular neuritis but usually less severe.

Shin and colleagues (2013) found that interictal Ménière's patients show milder vestibular SOT findings than acute vestibular neuritis patients, with substantial overlap between the two groups. CDP cannot reliably distinguish the diagnoses on pattern alone.

Some Ménière's patients have entirely normal interictal CDPs, especially in the early disease course. A normal interictal CDP does not exclude the diagnosis — clinical criteria and audiometry do that work.

Fluctuation and serial assessment

Reading levels: trainee · clinician · Anchor: menieres#fluctuation

The defining feature of Ménière's is fluctuation. Hearing fluctuates with attacks. Vestibular function fluctuates with attacks. CDP findings, measured serially, fluctuate too — typically worsening within a few days of an attack and improving over the subsequent weeks.

Serial CDP is therefore more informative than a single snapshot in this group. A patient with a normal CDP two weeks after their last attack tells you much less than the same patient with three CDPs over three months showing a pattern of worsening-then-recovery.

Late-stage Ménière's, after burnout of the affected labyrinth, typically shows a stable mild-to-moderate vestibular pattern. The fluctuation is gone because the residual function is no longer changing.

Archetypal signature

Interictal Ménière's typically shows a milder vestibular pattern than acute neuritis, and SOT findings fluctuate with disease activity (Shin et al. 2013). Repeated assessments over time provide more information than a single snapshot.

Signature jitter: 6 · MCT latency: 138 ms · ADT toes-up: [62, 50, 42, 36, 32]

BPPV and CDP

Reading levels: foundation · trainee · clinician · Anchor: bppv#overview

Benign paroxysmal positional vertigo presents with brief (typically less than 60 seconds) episodes of spinning vertigo provoked by specific head positions, most commonly rolling over in bed, looking up, or bending forward. The Dix-Hallpike manoeuvre reliably elicits the diagnostic nystagmus pattern.

BPPV is the most common cause of vertigo in clinical practice. Posterior canal involvement accounts for about 85% of cases; horizontal canal BPPV is less common but harder to recognise; anterior canal BPPV is rare.

CDP is essentially irrelevant to BPPV diagnosis. The test challenges quiet stance, not head position, and BPPV is provoked specifically by head movement. A normal CDP in a patient with positional vertigo does not exclude BPPV — the Dix-Hallpike does that work.

Findings between attacks

Reading levels: foundation · trainee · clinician · Anchor: bppv#between-attacks

Between BPPV attacks, CDP is typically normal or near-normal. The disorder doesn't affect the quiet-stance machinery; even during a symptomatic period, a patient tested when not actively dizzy will usually have a normal or near-normal SOT.

A patient with BPPV and an abnormal CDP usually has something else going on too — concomitant peripheral vestibulopathy, fear-avoidance behaviour, or an unrelated cause of imbalance. BPPV alone does not explain abnormal CDP findings.

This is why BPPV is the classic example of a vestibular diagnosis that requires the right test. Positional testing (Dix-Hallpike, supine roll) is diagnostic; CDP is unhelpful here.

After successful repositioning

Reading levels: trainee · clinician · Anchor: bppv#post-treatment

After a successful canalith-repositioning manoeuvre (Epley, Semont, BBQ roll depending on canal involved), the diagnostic positional nystagmus disappears and the patient's symptoms resolve. CDP, which was likely normal already, remains normal.

Persistent imbalance after successful BPPV treatment is not uncommon and is sometimes called 'residual dizziness' or 'sub-objective postural instability'. It typically resolves over days to weeks, may benefit from a brief course of vestibular rehabilitation, and rarely requires further investigation.

If imbalance persists for weeks after successful treatment, look for co-existing causes: PPPD overlay, concomitant peripheral vestibulopathy, or central pathology. A CDP at this point can be useful — not for the BPPV itself, but for whatever else might be contributing.

Archetypal signature

CDP is generally near-normal in BPPV between positional attacks — the disorder is provoked by specific head positions, not by quiet stance. A normal SOT does not exclude BPPV; positional testing (Dix-Hallpike, supine roll) is the diagnostic standard.

Signature jitter: 4 · MCT latency: 136 ms · ADT toes-up: [58, 46, 38, 32, 30]

Central vestibular disorders

Reading levels: foundation · trainee · clinician · Anchor: central#overview

Central vestibular disorders include strokes affecting the brainstem or cerebellum, demyelinating disease, cerebellar degeneration, brainstem tumours, and a long tail of less common pathologies. The common thread on CDP is that they affect the long-loop pathway, the adaptive machinery, or both.

Typical CDP findings in central disease are not subtle. The pattern is usually a combination of across-the-board SOT reduction (no specific input-isolated pattern), prolonged MCT latency, and impaired ADT adaptation. Any two of these three is highly suggestive; all three is essentially diagnostic of central involvement.

What CDP doesn't do is localise the central lesion. The same triad can arise from a cerebellar haemorrhage, a brainstem demyelinating plaque, or a chronic neurodegenerative process. Imaging and the rest of the neurological exam carry the localising work.

Cerebellar signs on CDP

Reading levels: trainee · clinician · Anchor: central#cerebellar

Cerebellar disease most often produces non-adapting ADT, sometimes with relatively preserved SOT and MCT. The ADT signature — sway energy on trial five roughly equal to trial one — reflects the cerebellum's role in motor learning.

Truncal ataxia from anterior-vermis lesions can produce broad-based SOT abnormality even on relatively easy conditions, sometimes with characteristically large mediolateral sway. Limb ataxia alone (interposed nuclei) is less likely to produce overt SOT abnormality.

Spinocerebellar degenerations follow this pattern with variable temporal evolution. SCA-3 (Machado-Joseph), SCA-6, and Friedreich ataxia all show progressive non-adapting ADT with gradually-evolving SOT abnormality; serial CDP can document the disease course objectively.

Brainstem involvement

Reading levels: trainee · clinician · Anchor: central#brainstem

Brainstem lesions affecting the descending vestibulospinal or reticulospinal tracts most often produce prolonged MCT latencies. The latency prolongation can be dramatic — 30–50 ms above normal in moderate-sized lesions.

Multiple sclerosis with brainstem demyelination is the most common cause of substantial MCT latency prolongation in younger patients. Lateral medullary (Wallenberg) stroke can produce striking CDP abnormalities ipsilesional to the stroke on some metrics.

Brainstem tumours producing CDP findings are usually large enough to be obvious on imaging by the time the CDP is abnormal. Subtle brainstem CDP findings without imaging correlates are uncommon and should prompt repeat imaging or a longer interval before diagnostic conclusions.

Differentiating central from peripheral

Reading levels: foundation · trainee · clinician · Anchor: central#diff-peripheral

The cleanest CDP discriminator between central and peripheral causes is the combination of SOT pattern and the other two protocols. A pure C5/C6 vestibular pattern with normal MCT and normal ADT adaptation is overwhelmingly peripheral. The same SOT pattern accompanied by prolonged MCT latency and non-adapting ADT is central until proven otherwise.

A few caveats. A normal head impulse test in a patient with abnormal CDP is a useful central pattern flag — peripheral lesions producing CDP findings usually have HIT abnormalities. Conversely, a positive HIT to one side with CDP findings matches the peripheral picture.

When the picture is mixed — for example, a peripheral SOT pattern with isolated ADT non-adaptation — the safest interpretation is 'mixed central-peripheral pathology' rather than pure either, with imaging to clarify the central component.

Archetypal signature

Central involvement is suggested by the triad of (1) abnormality across multiple SOT conditions rather than the selective C5/C6 pattern, (2) prolonged MCT latencies indicating long-loop pathway involvement, and (3) failure to adapt across ADT trials — the cerebellar signature.

Signature jitter: 6 · MCT latency: 175 ms · ADT toes-up: [82, 84, 80, 83, 81]

PPPD overview (Bárány Society criteria)

Reading levels: foundation · trainee · clinician · Anchor: pppd#overview

Persistent postural-perceptual dizziness (PPPD) is defined by Bárány Society 2017 criteria as three months or more of dizziness, unsteadiness, or non-spinning vertigo present on most days, exacerbated by upright posture, active or passive motion, and complex visual stimuli. The disorder is often triggered by a prior acute vestibular event (neuritis, BPPV, even a panic attack), and persists by maladaptive postural-control strategies.

PPPD is a clinical diagnosis. CDP, audiometry, MRI, and the rest of the workup are supportive — they document the absence of organic disease that would otherwise explain the symptoms — but they don't make the diagnosis. The criteria are symptom-based.

The disorder is common. Vestibular clinics commonly identify PPPD in 10–30% of chronic-dizziness referrals, depending on case mix. Many patients have been investigated extensively before the diagnosis is recognised; CDP plays a useful role in providing objective evidence consistent with the clinical picture.

Supportive CDP findings

Reading levels: trainee · clinician · Anchor: pppd#cdp-findings

The supportive CDP findings for PPPD are visual preference (elevated PREF ratio) and over-scaled MCT amplitudes. Either or both may be present; neither is required for the diagnosis.

Visual preference is the more specific finding — a PREF ratio above 1.0 in a patient with three months of dizziness and a normal structural workup is highly suggestive. The mechanism is hyper-reliance on visual cues: when vision becomes misleading (sway-referenced surround), the patient cannot disengage from it and sways with the false visual signal.

Over-scaled MCT amplitudes reflect anxious hypersensitivity. The patient over-corrects to small perturbations, producing larger sway than the perturbation warrants. This is non-specific (it can occur in any anxious patient) but supports the PPPD picture when present alongside visual preference.

PPPD vs aphysiologic pattern

Reading levels: clinician · Anchor: pppd#vs-aphysiologic

PPPD and aphysiologic patterns can co-exist, and distinguishing them requires care. PPPD is a clinical diagnosis with supportive CDP findings; aphysiologic patterns are CDP signatures that don't fit any physiologically-plausible disease. A patient can have both.

When the CDP shows clear visual preference plus over-scaling but no Cevette inversion, the picture fits PPPD without aphysiologic features. The patient is genuinely impaired by maladaptive postural control.

When the CDP shows a Cevette inversion or other aphysiologic features in addition to PPPD-supportive findings, the interpretation is more nuanced. Aphysiologic features can arise from extreme effort variability in anxious patients, or from a separate functional overlay. The clinical picture, the patient's effort during testing, and the consistency with other findings should drive interpretation rather than the CDP alone.

Archetypal signature

PPPD is a clinical diagnosis (Bárány Society criteria); CDP supports rather than confirms. The classic supporting feature is visual preference (worse on C3/C6 than C2/C5), reflecting hyper-reliance on vision. MCT amplitudes may be exaggerated.

Signature jitter: 6 · MCT latency: 138 ms · ADT toes-up: [68, 56, 46, 40, 36]

Aphysiologic patterns — definitions

Reading levels: foundation · trainee · clinician · Anchor: aphysiologic#overview

Aphysiologic patterns on CDP are configurations that don't fit any physiologically-plausible disease. The classic example is the Cevette inversion: condition 5 and 6 (the hardest, vestibular-only conditions) paradoxically better than condition 1 or 2 (the easiest). No real disease produces this.

Aphysiologic patterns are non-specific. They can occur in functional dizziness, deliberate effort modulation (malingering), severe anxiety, and occasionally in genuinely organic disease where extreme trial-to-trial variability produces atypical means. The pattern itself is a finding; the interpretation requires clinical context.

Reporting an aphysiologic pattern is appropriate when the configuration is clearly non-physiological. Naming it 'malingering' on the basis of CDP alone is not — many other causes can produce the same picture, and CDP is a poor lie-detector.

Cevette criteria

Reading levels: trainee · clinician · Anchor: aphysiologic#cevette-criteria

Cevette and colleagues (1995) defined a set of criteria for identifying aphysiologic CDP patterns. The criteria use multiple discordances: C5 better than C1, C5 better than C2, exaggerated inter-trial variability (typically more than 25 EQS points within a condition), and combinations of these.

The original criteria were derived in a sample including known feigners, validating against group-level discrimination. They have been used clinically since but have known false-positive rates in patients with severe anxiety and in some organic pathologies producing extreme variability.

A reasonable clinical threshold is: a single inversion (C5 > C1, for example) by more than 10 EQS points raises the question of aphysiologic pattern; multiple inversions or large inter-trial variability across multiple conditions strengthens the conclusion. Borderline configurations should be reported with appropriate uncertainty rather than over-called.

Differential considerations

Reading levels: trainee · clinician · Anchor: aphysiologic#differential

The differential for an aphysiologic CDP pattern includes:

Functional dizziness without intent to deceive — PPPD overlay, somatoform variants, and post-traumatic functional disorders can produce aphysiologic patterns through inconsistent effort or attention.

Severe anxiety — extreme anxiety during testing can produce trial-to-trial variability large enough to mimic aphysiologic patterns. Re-testing on a separate occasion in a calmer state can clarify.

Deliberate effort modulation — patients with secondary gain (litigation, disability claims) may modulate effort. The CDP cannot reliably distinguish this from genuine functional disorder.

Organic disease with extreme variability — rare, but possible in severe cerebellar disease, fluctuating Ménière's, and some demyelinating presentations. Cross-correlation with the other CDP protocols and the clinical picture is essential.

Medico-legal context

Reading levels: clinician · Anchor: aphysiologic#medico-legal

Aphysiologic CDP findings frequently surface in medico-legal contexts: personal-injury claims, disability evaluations, fitness-for-duty assessments. The temptation to use the finding as direct evidence of malingering should be resisted.

Appropriate medico-legal language describes what the CDP showed (an aphysiologic configuration meeting Cevette criteria), notes the differential (functional disorder, anxiety, effort modulation, atypical organic disease), and defers attribution to the multidisciplinary assessment. CDP is rarely the right test on which to base a definitive opinion about deception.

Where multiple tests across the work-up show consistent aphysiologic or non-organic patterns — for example, an aphysiologic CDP plus inconsistent posturography on different days plus give-way weakness on neurological exam plus a normal MRI — the cumulative weight may support a functional or feigned conclusion. Even then, the attribution belongs to the clinician integrating everything, not to the CDP report.

Archetypal signature

The Cevette (1995) hallmark: paradoxically worse performance on easy conditions (C1–C4) than on the harder C5/C6. Combined with high intertrial variability, this pattern raises the question of functional dizziness or feigned disease — but is non-specific and must be interpreted alongside the full clinical picture.

Signature jitter: 12 · MCT latency: 140 ms · ADT toes-up: [70, 65, 72, 60, 68]

Static posturography and centre-of-pressure metrics

Reading levels: foundation · trainee · clinician · Anchor: other-balance-tests#static

Static posturography records postural sway while the patient simply stands quietly on a fixed force plate — no moving platform or visual surround. It is far cheaper than computerised dynamic posturography and is widely available, including on low-cost and research-grade force plates.

The core output is the centre-of-pressure (COP) trajectory and the metrics derived from it: sway path length, the 95% confidence ellipse area, root-mean-square displacement, and mean sway velocity in the anteroposterior and mediolateral directions. Eyes-open and eyes-closed conditions are compared.

The eyes-closed-to-eyes-open ratio — the Romberg quotient — estimates visual dependence: a large increase in sway with eyes closed implies heavy reliance on vision to compensate for somatosensory or vestibular deficit. Static posturography localises poorly but is a sensitive, quantitative screen for instability.

The (modified) Clinical Test of Sensory Interaction on Balance

Reading levels: foundation · trainee · clinician · Anchor: other-balance-tests#ctsib

The Clinical Test of Sensory Interaction on Balance (CTSIB) is the bedside cousin of the SOT. The patient stands under four (or, in the modified mCTSIB, four) conditions crossing two surfaces (firm floor, foam pad) with two visual states (eyes open, eyes closed), and the examiner times how long balance is maintained.

Foam degrades somatosensory input the way the SOT's sway-referenced platform does; eyes closed removes vision. The eyes-closed-on-foam condition isolates vestibular-dependent balance — the bedside analogue of SOT conditions 5 and 6. Falling or stepping only in that condition points toward a vestibular contribution.

The mCTSIB requires only a foam pad and a stopwatch, costs almost nothing, and reproduces much of the SOT's sensory-conflict logic. It is the pragmatic first-line sensory-integration test where CDP is unavailable, and a reasonable triage to decide who needs formal CDP.

Functional balance and gait measures

Reading levels: trainee · clinician · Anchor: other-balance-tests#functional

Functional measures assess balance during whole-body tasks rather than quiet stance. The Berg Balance Scale, the Timed Up and Go, the Dynamic Gait Index, and the Functional Gait Assessment are validated, equipment-light tools that capture the real-world performance CDP cannot.

These scales predict falls, track rehabilitation, and translate directly into goals patients understand. Their weakness is that they do not separate the sensory contributions to imbalance — a low Berg score tells you the patient is unsteady, not why.

The complementary pairing is therefore CDP (or mCTSIB) to characterise the sensory mechanism, plus a functional scale to quantify the day-to-day impact and to anchor rehabilitation outcomes.

Choosing the right balance test

Reading levels: trainee · clinician · Anchor: other-balance-tests#choosing

Match the test to the question. To document the functional impact of a known vestibular deficit and to separate sensory contributions, CDP or mCTSIB earns its place. To screen broadly and cheaply, static posturography and functional scales suffice.

Availability, cost, and the patient's tolerance all bear on the choice — frail or acutely vertiginous patients may manage a Timed Up and Go but not several minutes of sway-referenced stance. No single test is comprehensive; the balance battery is deliberately layered.

Critically, none of these tests localises a lesion. They quantify function, and are interpreted alongside the structural vestibular work-up — calorics, vHIT, VEMPs, and imaging — that answers the where question.

Wearable inertial sensors (IMUs)

Reading levels: foundation · trainee · clinician · Anchor: emerging-tech#wearables

Wearable inertial measurement units (IMUs) — the accelerometers and gyroscopes already inside every smartphone — can quantify postural sway and gait from a sensor clipped to the lower back or worn on the limbs. They are inexpensive, portable, and require no fixed laboratory.

IMU-derived sway metrics correlate reasonably with force-plate measures, and IMUs add gait and turning analysis that fixed plates cannot capture. The trade-off is lower spatial precision and sensitivity to placement and calibration, so they complement rather than replace laboratory posturography.

Their real promise is reach: balance assessment that can happen in a clinic corridor, a care home, or a patient's living room, repeated as often as needed, widening access well beyond the few centres that own a CDP platform.

Virtual reality and immersive posturography

Reading levels: foundation · trainee · clinician · Anchor: emerging-tech#vr

Virtual-reality headsets recreate the moving visual surround of CDP at a fraction of the cost, and go further — immersive scenes can deliver richer, more naturalistic optic-flow conflict than a physically tilting surround ever could.

Combined with a force plate or IMU, VR enables sensory-conflict paradigms, graded exposure for visually-induced dizziness, and gamified vestibular rehabilitation that improves engagement and adherence. Early studies in PPPD and visual vertigo are encouraging.

Caveats remain: cybersickness, the absence of agreed normative databases, and headset-to-headset variability. VR posturography is a fast-moving research and rehabilitation tool rather than a standardised diagnostic test today.

Machine learning and automated interpretation

Reading levels: trainee · clinician · Anchor: emerging-tech#ai

Machine-learning models trained on posturographic and CDP data can classify disease patterns, flag aphysiologic results, and predict fall risk, sometimes matching or exceeding rule-based interpretation on the datasets they are trained on.

The barriers are familiar ones: modest, single-centre datasets; weak external validation; device-specific signals that do not transfer; and the opacity of models whose reasoning a clinician cannot inspect. Regulatory and medico-legal frameworks for automated balance interpretation are still immature.

The realistic near-term role is augmentation — automated artefact detection, normative comparison, and triage that surfaces the cases needing expert review — rather than autonomous diagnosis. The clinician remains responsible for the interpretation.

Home monitoring, telehealth, and the road ahead

Reading levels: trainee · clinician · Anchor: emerging-tech#future

The convergence of cheap IMUs, VR, and machine learning points toward balance assessment that is continuous, home-based, and connected — periodic sway and gait sampling that detects deterioration between clinic visits and triggers earlier intervention.

For that future to arrive responsibly the field needs shared normative databases, cross-device standardisation, and prospective validation against hard outcomes such as falls — the same evidential rigour that underpins established CDP.

Used well, these technologies extend the reach of balance-function testing far beyond the specialist laboratory. Used uncritically, they generate impressive-looking numbers with no validated meaning. The discipline of pattern recognition and clinical correlation that the rest of this atlas teaches applies just as firmly to the new tools.

What CDP changes in clinical practice

Reading levels: foundation · trainee · clinician · Anchor: clinical-applications#overview

CDP rarely makes a diagnosis on its own. Its clinical value is to change management decisions — to confirm a functional deficit when structural tests are equivocal, to quantify impact, to stratify fall risk, and to track change over time.

The recurring question CDP answers is how is this patient coping, given whatever organic substrate the rest of the work-up has revealed. Two patients with identical calorics can have very different stability profiles, and CDP is what distinguishes them.

The applications below are organised by clinical domain. For each, the useful pattern, the decision it supports, and the strength of the underlying evidence are what matter — not the raw scores in isolation.

Application domains — from diagnosis to monitoring

Reading levels: foundation · trainee · clinician · Anchor: clinical-applications#domains

Peripheral vestibular. Disproportionately low SOT 5/6 with preserved 1–4 supports a vestibular contribution to chronic imbalance when caloric or vHIT data are borderline, and biases toward a gaze-stabilisation and substitution rehabilitation programme.

Central nervous system. Prolonged MCT latencies, failure of ADT adaptation, and inconsistent or multi-system patterns raise suspicion of central involvement — cerebellar disease, Parkinsonism, or demyelination — and prompt imaging and neurology referral.

Rehabilitation, fall risk, concussion, and the functional/non-organic domain. Serial composites document VRT progress; constricted limits of stability and surface/visual dependence flag fall risk; post-concussion CDP tracks recovery; and aphysiologic patterns provide objective, non-accusatory evidence in medico-legal settings.

Fall-risk stratification

Reading levels: trainee · clinician · Anchor: clinical-applications#fall-risk

A low SOT composite, falls on conditions 5 and 6, combined surface-and-vision dependence, and a slow, constricted limits-of-stability envelope together identify patients at elevated risk of falling — particularly older adults and those with bilateral vestibular loss.

CDP-based fall-risk stratification is most useful when it changes the plan: triggering a home-safety assessment, a targeted balance-training referral, medication review, and patient education about high-risk situations such as walking in the dark.

It is one input among several. Functional scales, medication burden, vision, cognition, and environmental hazards all feed the overall fall-risk picture, and CDP should be read in that wider context.

Guiding and monitoring rehabilitation

Reading levels: trainee · clinician · Anchor: clinical-applications#rehab

CDP helps design vestibular rehabilitation by revealing which sensory channel the patient over- or under-relies on. Heavy visual dependence argues for optokinetic and visual-conflict exposure; surface dependence argues for compliant-surface and proprioceptive training.

Serial testing then documents change objectively. Improving SOT 5/6 scores and an expanding limits-of-stability envelope provide motivating, defensible evidence of progress that symptom diaries alone cannot supply.

Care is needed to distinguish genuine improvement from a learning effect — patients get better at the test itself. Spacing retests, interpreting trends rather than single sessions, and pairing CDP with a functional scale guard against over-reading practice gains.

Evidence and limitations

Reading levels: clinician · Anchor: clinical-applications#evidence

The evidence base is strongest for CDP as a functional and monitoring tool and weakest for any claim of diagnostic specificity. Patterns are supportive, never pathognomonic — the same SOT profile arises from neuritis, schwannoma, ototoxicity, or a brainstem lacune.

Bárány Society consensus documents reference CDP findings as supportive evidence for several disorders but never as primary diagnostic criteria, and that posture is appropriate to the data.

Used as a measure of function and change, interpreted alongside the structural work-up and the patient in front of you, CDP earns its place. Used as a standalone diagnostic oracle, it overpromises. The discipline is to let it answer the functional question it is good at.

Quick-reference tables

Reading levels: foundation · trainee · clinician · Anchor: tools#reference-tables

Quick-reference summaries are gathered here for clinical use. Tables and figures throughout the atlas can be printed via the print buttons on each module page, or the entire atlas can be exported via /print-all.

Latency norms (MCT): 120–150 ms in healthy adults under 60. Upper limit of normal ~165 ms. Latencies above 170 ms warrant attention to central pathology.

Composite norms (SOT): ≥70 in healthy adults under 60; ≥60 by age 75. Device-specific norms apply.

Sensory ratio cutoffs: SOM ≥0.90, VIS ≥0.75, VEST ≥0.60, PREF ≤0.92. Below these values is the abnormal direction (above for PREF).

Disease pattern matrix: Vestibular pattern → peripheral (neuritis, Ménière's, schwannoma). Across-the-board reduction + prolonged latency + non-adapting → central. Visual preference + over-scaling → PPPD. Cevette inversion → aphysiologic (consider functional / effort modulation).

Reporting checklists

Reading levels: trainee · clinician · Anchor: tools#checklists

A reporting checklist for a complete CDP report:

1. Test conditions — date, device, examiner, patient height and stocking-feet stance, harness adjustment, calibration status.

2. SOT — composite score, condition-by-condition means, falls per condition, sensory ratios with classification (normal / reduced / markedly reduced).

3. MCT — latencies for small / medium / large translations in each direction; amplitude scaling; weight symmetry.

4. ADT — sway energies trial-by-trial in each direction; adaptation present / absent.

5. Pattern recognition — name the canonical pattern (vestibular / visual-vestibular / surface-dependent / visual preference / central triad / aphysiologic / mixed).

6. Clinical correlation — relevant prior findings (audiometry, HIT, calorics, imaging), the question the test was asked to answer, and the answer given.

7. Caveats and limitations — anything that compromises the test (harness assist, anxious effort, atypical configurations).

Calculators

Reading levels: foundation · trainee · clinician · Anchor: tools#calculators

Two simple calculations are useful in CDP interpretation but rarely shown explicitly on commercial reports.

Falls-weighted composite: a single fall in three trials drops the condition mean by 33 points if scored as zero. Some clinicians prefer to report fall trials separately so the underlying sway profile is visible.

Age-adjusted composite: a rough adjustment subtracts 0.4 points per year over age 60. A 75-year-old patient with a composite of 64 is borderline normal; the same composite in a 30-year-old is clearly abnormal. Device-supplied age-norm tables are more accurate.

These calculators are not built into the atlas as interactive tools — the values are simple enough that explicit computation rarely changes a clinical decision. The pattern recognition layer carries the diagnostic information.