BPPV & the Epley Maneuver: How Loose Ear Crystals Spin the World

The normal inner ear: two sensors, one shared physics

To see what breaks in BPPV, start with what works. Each inner ear contains a vestibular labyrinth with two kinds of motion sensor. The otolith organs — the utricle and saccule — sense linear acceleration and the pull of gravity. Their sensory surface, the macula, is covered by a gelatinous membrane studded with thousands of tiny calcium-carbonate crystals called otoconia (or otoliths), each roughly 3–30 µm across. Because these crystals are denser than the surrounding fluid, gravity tugs them, shearing the underlying hair cells and telling the brain which way is down.

The second sensor type is the three semicircular canals — anterior, posterior, and horizontal (lateral) — arranged at roughly right angles so that between them they detect rotation in any plane. Each canal is a fluid-filled ring. At one end sits the ampulla, containing a gelatinous flap, the cupula, draped over hair cells. When you rotate your head, inertia makes the fluid (endolymph) lag behind, deflecting the cupula and firing the hair cells. Crucially, in health the cupula has the same density as endolymph, so it is neutrally buoyant and completely insensitive to gravity. The canals report rotation; the otoliths report gravity and linear motion. That clean division of labor is the key to everything that follows.

What goes wrong: crystals in the wrong compartment

BPPV begins when otoconia detach from the utricular macula. This happens with age (the gel degenerates), after head trauma, after prolonged bed rest, or often for no identifiable reason. The freed crystals migrate — and because of anatomy and gravity, they most often fall into the posterior semicircular canal, which sits in the most dependent (lowest) position when a person is upright or lying down. This accounts for roughly 85–95% of cases; the horizontal canal is next (~5–15%), and the anterior canal is rare (<1–2%).

The dominant mechanism is canalithiasis — crystals floating free within the canal fluid. The causal chain is elegant and entirely mechanical:

1. The patient moves the head into a provoking position (rolling over in bed, tipping the head back, looking up).
2. The now gravity-dense crystal debris begins to sink to the lowest point of the canal.
3. As the plug of crystals moves, it drags endolymph with it, like a piston in a tube.
4. That fluid current deflects the cupula — even though the head has stopped moving.
5. The brain receives a powerful, sustained rotation signal from one canal that is contradicted by the eyes and the opposite ear.

The mismatch is interpreted as spinning. A less common variant, cupulolithiasis, occurs when crystals adhere directly to the cupula, making it heavy and gravity-sensitive; this tends to cause a longer-lasting nystagmus.

Why it spins, why it's brief, and why the eyes betray it

Two hallmark features of BPPV fall directly out of the canalithiasis model. First, the latency: after the head is positioned, there is a delay of about 1–5 seconds before vertigo begins, because the crystals must overcome inertia and start to sink. Second, the vertigo is brief and self-limiting — usually under 60 seconds — because once the crystals reach the bottom of the canal and stop moving, the endolymph current ceases, the cupula springs back to neutral, and the false rotation signal ends.

The eyes give the diagnosis away. Because the vestibular system drives the vestibulo-ocular reflex (VOR), a false rotation signal produces involuntary eye movements — nystagmus. Stimulation of the posterior canal generates a characteristic torsional (rotatory) and upbeating nystagmus, with the top poles of the eyes beating toward the affected (lower) ear. This nystagmus fatigues on repeated testing, as the crystals disperse. The tight coupling between a specific canal's geometry and a specific eye-movement pattern is why an examiner can often name the exact canal and side purely from watching the eyes.

Clinical presentation and the Dix-Hallpike diagnosis

Epidemiology. BPPV is the single most common cause of vertigo, with a lifetime prevalence around 2–3% and an estimated annual incidence of roughly 0.6% of the population. Incidence rises sharply with age; the peak is 50–70 years, and women are affected about 2–3× more often than men.

Typical presentation. Patients describe sudden, brief spinning triggered by specific movements: rolling over in bed, lying down, sitting up, tipping the head back to a high shelf, or bending forward. Episodes last seconds, not hours. Between episodes patients may feel a lingering unsteadiness or nausea, but there is no hearing loss, no tinnitus, and no ear fullness — their absence is an important clue.

Diagnosis is clinical, not radiological. The confirmatory test for posterior-canal BPPV is the Dix-Hallpike maneuver: the seated patient's head is turned 45° toward the tested ear, then the patient is laid back quickly so the head hangs ~20–30° below horizontal. A positive test reproduces the vertigo with the classic latent, torsional-upbeating, fatigable nystagmus. For suspected horizontal-canal BPPV, the supine roll (Pagnini-McClure) test is used instead. Imaging (MRI) is not needed for typical BPPV — it is reserved for atypical features that suggest a central cause.

The Epley maneuver: repositioning by gravity

The Epley canalith repositioning maneuver, introduced by John Epley in 1992, treats posterior-canal BPPV by using gravity to walk the crystal debris out of the canal and back into the utricle, where it can no longer generate false rotation signals. It is a sequence of four to five held head positions, each rotating the plane of the posterior canal so the debris slides forward toward the canal's exit (the common crus) rather than back toward the ampulla. A representative sequence (for a right-sided lesion):

1. Seated, head turned 45° to the affected (right) side.
2. Lay supine quickly with head extended ~20° over the table edge (this is the Dix-Hallpike position) — hold ~30–60 s until nystagmus subsides.
3. Rotate the head 90° to the opposite (left) side — hold ~30–60 s.
4. Roll the whole body onto the left shoulder and turn the head a further 90° so the nose points toward the floor — hold ~30–60 s.
5. Sit up slowly with the chin slightly tucked.

A single treatment resolves symptoms in roughly 70–90% of patients; a second cycle raises success further. It is the mechanistic elegance that makes BPPV so satisfying to treat: the problem is a misplaced object, and the cure is simply putting it back.

Red flags, natural history, and a common misconception

Natural history. Untreated BPPV often remits spontaneously over weeks to months as crystals dissolve or disperse, but it frequently recurs — roughly 15–50% within a few years. Its real morbidity is indirect: in older adults, positional vertigo is a significant, under-recognized contributor to falls and fractures, so treating it is more than symptom relief.

Red flags that argue against BPPV. Vertigo that is constant rather than positional, direction-changing or purely vertical (down-beat) nystagmus, vertigo with a new severe headache, or any accompanying neurological deficit — diplopia, dysarthria, dysphagia, limb weakness, or truncal ataxia — should raise concern for a central cause such as a cerebellar or brainstem stroke. The mnemonic to remember is that isolated, well-behaved positional vertigo is reassuring; positional vertigo with company is not.

The common misconception. Many patients and even clinicians assume vertigo means something is wrong with the brain, or that a spinning sensation warrants an urgent CT scan. In fact CT is poor for the posterior fossa and normal in BPPV; the diagnosis is made by watching the eyes during positioning, not by imaging. A second frequent error is confusing vertigo (an illusion of motion) with lightheadedness or presyncope — the latter points to blood pressure, the heart, or medications, not the labyrinth. Getting that first fork right saves enormous unnecessary testing.

BPPV vs. its most dangerous mimic — distinguishing peripheral positional vertigo from a central (brainstem/cerebellar) cause.

Feature	BPPV (peripheral)	Central positional vertigo (red flag)
Latency after positioning	Brief delay ~1–5 s, then vertigo	Often no latency — immediate
Duration of episode	Seconds, typically <60 s, self-limits	Persistent, may not fatigue
Nystagmus direction	Torsional-upbeat, toward down ear (posterior canal)	Pure vertical/down-beat, direction-changing
Fatigability on repetition	Yes — diminishes with repeated testing	No — reproduces each time
Other neuro signs	None (isolated vertigo)	Diplopia, dysarthria, ataxia, limb signs