Tullio's Phenomenon

Timothy C. Hain, MD •Content last updated: March 27, 2021

Types of Fistulae
Locations in the inner ear associated with fistula.



The "Tullio phenomenon; consists of dizziness induced by sound. For example, use of one's own voice or a musical instrument.

Tullio's occurs mainly in five ear conditions: Superior canal dehiscence, perilymph fistula, Meniere's syndrome, post fenestration surgery, and vestibulofibrosis. Vestibular atelectasis is also a potential cause.

The most common cause at this writing is superior canal dehiscence.

Meniere's syndrome is the second common because of its relatively high prevalence (about 2/1000). Vestibulofibrosis, or the attachment of a motion sensitive part of the ear to the stapes, is very uncommon.Perilymph fistula, is probably the next to last, and fenestration the least common cause. There are only a very few people still living who have had fenestration surgery for otosclerosis as this operation has been out of favor for many years. Accurate figures regarding the prevalence of fistula and vestibulofibrosis are presently not available.

Tullio's is probably more common in persons who wear hearing aids, because there is more sound energy, although this has never been studied.

Supplemental material:

Supplemental material: Movie of nystagmus elicited by pressure

Tullio's often is accompanied by hearing loss, tinnitus, and pressure sensitivity. The latter symptom is rare, but arises from similar mechanisms and disease processes as does sound sensitivity. Sensitivity to Valsalva is another type of pressure sensitivity.

Supplemental material: Movie of nystagmus elicited by Valsalva


There are probably several mechanisms for Tullio's phenomenon as several types of disease processes have been associated with Tullio's.

Tullio (1929) originally investigated sound-induced eye and head movement in pigeons. He made openings in the semicircular canals and demonstrated that sound waves spread primarily into the canals that had been opened. Huizinga (1935) proposed that the openings allow sound energy to move the ampulla of the semicircular canal, and furthermore, that the net effect is greater in the direction that excites the vestibular nerve. This is similar to Ewald's second law. Rottach and associates, implicated stimulation of the horizontal canal (1996). Persons with Perilymph Fistula are thought to have Tullio's related to an abnormal opening in the inner ear, at one of the weak points shown in the first figure above.

Nadol suggested that the stapes foot plate may form fibrous adhesions to the utricle spontaneously or as a post infectious process, terming this process "vestibulofibrosis".

Nadol also suggested that pathologic dilation of the otolithic sac, as in Meniere's disease, might cause it to become more sensitive to movement of the stapes.

Finally, collapse of perilymphatic membranes, called "vestibular atelectasis" or "floating labyrinth", might allow it to become a force conduit between the stapes and the utricle. This is currently (as of 2019) an obscure and controversial idea.

With respect to mechanisms involving abnormal stapes movement, Dieterich (1989) suggested that a hypermobile stapes with annular ligament damage or subluxation of the stapes may cause abnormal utricular contact. Combining both ideas, Kwee (1976) postulated that a congenital middle ear bone malformations may predispose the utricle to abnormal fluid movement generated by an irregularly shaped stapes.

Fenestration surgery, or in other words, the making of openings in canals, was used for otosclerosis during the 1940's, before the invention of the stapes operation for otosclerosis. As of 2019, there are very few patients who have fenestrations, as the operation as well as the patients have died out. Nearly all of these patients have Tullio's.


Figure 2. Coronal thin cut CT scan showing superior canal dehiscence (SCD). This patient was reported in detail in (Ostrowski, Hain and Wiet, 1997).

See also Supplemental material : Movie of nystagmus elicited by sound

Superior canal dehiscence (SCD), a thinning or absence of the roof over the superior semicircular canal has recently been described. Approximately 2% of the population has SCD on autopsy. SCD is associated with movements in the plane of the superior canal on stimulation. SCD is also associated with low-threshold sound evoked vestibulocollic reflexes, also known as VEMP's (Brantberg et al, 1999; Watson et al, 2000). Perilymph fistula can overlap with SCD and Meniere's disease, but Tullio's is uncommon in Meniere's disease.

Hyperekplexia is a well known central neurologic disorder that has some similarities to Tullio's (Zhou et al, 2002). Persons with this disorder are abnormally sensitive to sound.


Tullio's is a symptom, not a disease, so you diagnose it by simply making the observation that loud noises make one dizzy. We don't mean that loud noses make you uncomfortable -- chalk screeching on the blackboard and all that, but literally dizzy as in spinning vertigo. It is particularly important to try to determine which ear is the problem. The next step should be to consult with a doctor who knows something about this. Generally this will be an otolaryngologist who specializes in ear problems (an otologist). This doctor will likely do a hearing test, take acareful history, and get a temporal bone CT scan to look for superior canal dehiscence and other ear problems.

The Tullio test -- an analysis of the methodology.

Occasionally doctors will attempt a formal "Tullio's" test. In our practice in Chicago, we do this ourselves (in persons who complain of sound sensitivity), and think that this is rarely positive, but still worth the minimal effort that it takes to assess Tullio's. There are several methods of doing this.

From a systems perspective, there are three potential variables in testing for tullio's:

Tullio test: Sound generators

Starting with sound generators, audiometers (hearing testing machines) are the most accessible to most clinicians. Here, Tullio's is noticed serendipitously by an observant audiologist. Because Tullio's is generally not elicited by soft sounds, the pitfall of this method is that in most persons, the stimulus is not loud enough to cause any problem. Another pitfall is that only one frequency is tested at a time. One might miss the sound that triggers symptoms.

Barany Noise Box used for Tullio test Shower massager used for Tullio test


The "Barany Noise Box" -- this is a wind-up device with a nozzle that one puts into the patient's ear. This is a reasonable approach -- it is no longer available through surgical medical suppliers -- we are not sure why. We have not compared it to our usual method (see below).

In the author's clinical practice in Chicago, a simple alternative device is used --- a shower massager. This inexpensive device, also useful for the extremely useful vibration test, produces a loud low-pitched buzzing noise that can provoke dizziness and nystagmus in persons with Tullio's, when put close to the ear for a second. It is certainly possible that a device like the Barany noise box might be more suitable. In our informal testing, the Barany noise box seems to be less well tolerated by patients, perhaps because it is louder or because it has a different spectrum of frequencies.

Note that this device (shower massager) and similar ones are not approved "medical" devices -- the noise level that it produces is uncalibrated, and theoretically, it could be injurious to hearing. We do not recommend that you use this type of device or other sound generator unless you have calibrated it are certain that it does not produce injurious noise.

The movie shown above of a subtle nystagmus elicited by sound uses this method.

Tullio-test: Intrinsic variability

Moving next to person's sensitivity, this varies immensely. While the literature mainly concerns people with profound sound sensitivity, in reality, there is a broad spectrum of sensitivity. Persons with Meniere's disease and window fistulae, generally have tiny responses (if any) to loud noise. So they are at the low end of the spectrum. Persons with fistulae often have fatigable Tullio's (and responses to pressure). Persons with disorders due to dilation or adhesions of the inner ear (e.g. Meniere's, vestibulofibrosis), and superior canal dehiscence (SCD) have reproducible Tullio's.

Persons with superior canal dehiscence, may have very large responses to sound, but again they are variable - they may have small or large openings in their superior canal, they may have bilateral dehiscence.

Tullio-test: Output measurement

The biggest variable of all is output measurement. The most common method is simply to ask the patient if they are dizzy, or perhaps to watch their eyes. Subjective measures are sensitive but are vulnerable to observer bias. Just "eyeballing" the patient's eyes is very insensitive -- only gigantic Tullio's will be seen in this way (as we show on some of our movies).

Video Frenzel Goggle system

We think that the best device is a video-frenzel goggle system (shown above). The head must be held by the examiner when using this, as otherwise, people tend to jump with loud noises.

ENG systems are generally terrible for Tullio's -- they combine low resolution with a lack of an ability to know when the sound was generated. They also are unable to measure torsion - -which is usually the main response to sound.

This is an example of the unusual situation right now (4/2007) where office-based diagnostic techniques for dizziness have greatly outstripped commercial testing systems.

Microscope method: If there are two people available, and an examining microscope, the microscope can be used to watch scleral blood vessels while another person produces the noise with some device. Again the head must be immobilized. While very good, this method is generally impractical as it depends on having two examiners present at the same time.


Unusual and/or promising ideas that so far have not been established

Ben-David et al (1997) reported that posturography could be used to diagnose Tullio. This seems worth pursuing. Logically it would seem that loud noises might also trigger postural readjustments.

It would seem likely that Tullio's would affect gait -- here one would need to assess gait with and without loud noise -- perhaps walking with eyes closed while listening to loud music. Of course, this is close to the situation that patients complain about.

Response triggered averaging could potentially make ENG testing superior to office-based Tullio testing. For this to happen, we will need some movement by ENG manufacturers to improve their equipment.

Treatment of Tullios.

Treatment is based on the cause. Generally it is not terribly effective. Tullio's is generally an annoyance. It is not usually rational to undergo risky operations to fix an annoyance.

Nevertheless, SCD can be roofed or plugged. Fistulae can be patched, Meniere's can be treated medically and surgically. Fenestrations may be closed, although this is done very rarely.

Vestibulofibrosis can be treated with approaches that destroy vestibular function, disconnecting the ear from the brain.

More practically, you may find it useful to wear ear plugs, and avoid loud noises.