Timothy C. Hain, MD
Page last modified: September 23, 2021

The main BPPV page is here.

Ear Rocks

There are several variants of BPPV (Benign Paroxysmal Positional Vertigo) which may occur spontaneously as well as after the Brandt-Daroff maneuvers or Epley/Semont maneuvers. They are thought to be caused by migration of otoconial debris into canals other than the posterior canal, such as the anterior or lateral canals. It is also theoretically possible for many aberrant patterns of BPPV to occur from an interaction of debris in several canals, location of debris within the canal, and central adaptation patterns to lesions. For this reason, in clinical practice, atypical BPPV is first treated with maneuvers as is typical BPPV, and the logic outlined below is entered into only after treatment failure.

Lateral canal BPPV is the most common atypical BPPV variant, accounting for about 3-12 percent of cases (Cakir et al, 2006; Korres et al, 2002; Hornibrook, 2004). It seem likely that the percentage of lateral canal BPPV depends greatly on the criteria one uses -- allowing diagnosis from smaller amounts of nystagmus is associated with larger prevalence. In the author's experience, most lateral canal cases are seen as a consequence of an Epley maneuver, but others find that spontaneous occurence is more common (Hornibrook, 2004). Diagnosis requires seeing a horizontal nystagmus that changes direction depending on the down ear. In the post-Epley situation, it is generally very strong.

Looking at the diagram above, it is hard to see how anyone could have persistent lateral canal BPPV, because the lateral canal is tilted so that debris should roll down the canal into the vestibule. Just walking around should treat it. By the same token, it is difficult to see how lateral canal BPPV could persist overnight, given that most people roll from one side to the other in bed. It would seem likely from this logic that the simple anatomy of the lateral canal geometry depicted above is not entirely accurate.

Causes of Lateral canal BPPV.

Canalithiasis of lateral canal:

Displaced otoconia can migrate to the posterior canal, which is the lowest part of the ear when one's head is upright. Debris can also migrate into the lateral canal as well as the superior canal. For the superior canal, debris would tend to fall out spontaneously unless it was at the ampulated end. For the lateral canal, debris also would tend to fall out spontaneously. This may be the reason that non-posterior canal BPPV is unusual.

Cupulolithiasis, and light cochlea:

Debris can not only migrate into the long arms of the canals, but might also become adherent to the cupulae. This is called "cupulolithiasis". Injuries to the cupula such as due to infection or poor circulation can also, in theory, cause cupulolithiasis. These are examples of "heavy cupula", and cause an ageotrophic nystagmus.

Other substances in the inner ear might also cause a nystagmus resembling lateral canal BPPV. Positional alcohol nystagmus is well known. In theory, protein in the inner ear such as from an intracochlear schwannoma might result in a "light cupula" situation, and a geotrophic nystagmus.


Nystagmus of other disorders such as central positional nystagmus due to cerebellar damage can closely resemble lateral canal BPPV (Lee et al, 2014), but fortunately are much less common. With very sensitive devices to look for nystagmus, it is far more common to see horizontal direction changing nystagmus, that might be mild lateral canal BPPV.

Diagnosis of lateral canal BPPV

Patients with lateral canal BPPV are usually very dizzy with their head turned to either side in bed. This is very different than the situation with posterior canal BPPVwhere one is dizzy only to the "bad side". In anterior canal BPPV, symptoms are often worse straight back.

Practically, lateral canal BPPV can almost always be seen on the Dix Hallpike test, especially if the examiner does not attain a substantial head-hanging posture but instead tests the patient supine. Nevertheless, the best position to see the direction changing horizontal nystagmus of lateral canal BPPV is not the Dix-Hallpike maneuver. Rather one starts with the body supine, head inclined forward 30 degrees, and then turns the head 90 deg to either side. This is called the "supine roll test". The logic of the supine roll test for lateral canal BPPV is that the 30 degree inclined forward position puts the planes of the lateral canals approximately in the horizontal plane, and thus turning the 90 degrees head to either side causes a larger change in the force on the otoconia. The optimal amount to turn the head for lateral canal is 90 degrees, rather than the 45 degrees which is optimal for PC BPPV.

.Supplemental material

The nystagmus of lateral canal BPPV can be either always towards the ground ("geotropic") or always towards the sky ("ageotropic", or "apogeotropic" -- we will use the shorter construction)(Bertholon et al, 2002). Nystagmus that is "ageotropic" (about 25%) is thought to be caused by debris that is further around the canal and closer to the ampulla, than "geotropic" nystagmus (about 75%). It is unlikely that debris is actually adherent to the cupula as this should not cause much vertigo (Hain et al, 2005).

Lateral canal BPPV can cause a very strong and prolonged vertigo. People with lateral canal BPPV are also generally more disturbed by ordinary sideways rotational head-movements than people with posterior canal BPPV. Lateral canal BPPV may occur commonly but may also be self treated as people roll back and forth at night naturally during sleep (Korres et al, 2002).

In some cases, usually ones where the condition occurs spontaneously rather than as a consequence of treatment for regular BPPV, debris is adherent to the cupula. This causes a very prolonged and refractory nystagmus, but it is rarely very strong.

Misdiagnosis is possible : One should be more concerned in lateral canal BPPV, than in PC BPPV, that there is a cause other than BPPV. Disorders that resemble BPPV, including (rarely) brain tumors, are discussed here. Vestibular migraine is certainly possible (Beh, 2018).

Which is the bad ear ?

When lateral canal BPPV follows a treatment maneuver for posterior canal BPPV, the "bad" ear is considered to be the same one with the posterior canal BPPV.

In idiopathic cases with geotropic nystagmus the "bad" ear is assigned to the side with the stronger nystagmus. With ageotropic nystagmus, the bad ear is assigned to the side with the weaker nystagmus. The rationale for this pattern is that excitation is stronger than inhibition (i.e. Ewald's second law). This mechanism was not supported by a recent study of positional alcohol nystagmus on persons who had only one remaining labyrinth (Tomanovic and Bergenius, 2013).

As the lateral canal, when upright, is mainly horizontal, debris might in theory be distributed nearly anywhere in the lateral canal. Then when one lies down, the nystagmus that appears might be dependent on the initial distribution.

Han et al (2006) suggested that the nystagmus seen on lying supine is can be used to determine which ear is affected. The methodology here is that the patient is initially sitting with head bent down for 3 minutes, and then rapidly brought into the supine position, with the head on a pillow. For geotropic nystagmus, nystagmus is away from the affected ear, and for ageotropic, towards the affected ear. In other words, for geotrophic nystagmus, the nystagmus follows the general rules for paretic ears, and vice-versi for ageotrophic nystagmus.

Methods that we are dubious about:

Some authors feel that a "null point" can be found, and the bad ear is on the side of the null point (Bisdorff and Debatisse, 2001). We find this generally implausible as if debris is movable within the canal, it should always die away, and thus there is no null point. If useful, it should only apply to cupulolithiasis, which is very rare.

Bottom line regarding localization:

Usually there will need to be judgment call on the part of the examiner, integrating together information about nystagmus and other data about which ear is diseased (such as hearing, fullness and the like). When horizontal nystagmus follows an Epley maneuver for posterior canal BPPV, in nearly all instances the most likely "bad" ear is the one in which posterior canal type BPPV was seen previously. In situations where the side is unclear, it is our practice to first treat the more likely side with home exercises, and then switch to the other side after a week.

Treatment of lateral canal BPPV

Treatment of lateral canal BPPV has not been as well established as in typical PC BPPV, but it is based on the same biomechanics and logic.

All maneuvers for lateral canal BPPV take the general approach of turning the body or head around the long axis, from the "affected" side, towards the good side. Variant maneuvers can all be viewed as "pieces" of the log-roll. Thus, if you understand the log-roll, unless you are very interested in saving time by using a variant maneuver, you don't have to read beyond the log roll section. Of the variant maneuvers, the "Zuma" is the closest to the log roll. Is it better ? We don't know.

Oron et al (2015) reviewed numerous maneuvers for lateral canal BPPV. While this certainly was not an exhaustive review, and it was unaccompanied by any biophysical simulations, they stated " No treatment was found to be superior over the others regarding the success rate." Our thought is that in as much as one often doesn't know the bad side, the log roll is a simple and effective treatment.

Log Roll (also known as barbecue, or "Lempert")

Log roll exercises for Lateral Canal BPPV (c) Timothy C. Hain, M.D. 2013

The "log roll" exercises, are a procedure where an individual is rolled in steps of 90 deg, starting supine/affected ear down, to supine, to affected ear up, to nose-down, and then to sitting at intervals of 30 seconds or one minute.  This procedure seems very reasonable and it is the one that we often use in our own clinical practice. There is a report of 75% efficacy (15/20)  of a variant procedure (e.g. Fife, 1998) called the "iterative full-contralateral roll",  going from supine nose up, a full 360 degrees in 90 degree increments, rotating towards the good ear. This procedure is performed once or twice in the clinic and repeated at home for 7 days. It seems to us that the difficulty of establishing which is the "bad" ear is an obvious drawback of almost any lateral canal treatment and in some situations, we do the log roll to one side for a week, and follow with the log roll to the other side for another week. We also feel that it is preferable to begin with the bad-ear down rather than supine, for situations where there is debris close to the ampula (i.e ageotrophic). Vibration of the mastoid might theoretically add to efficacy of this procedure but no studies are available at the present writing.

Right now, although position 4 says "hands/Knees", actually we just have people roll over to nose down and cradle their head in their hands. This attains the same position without as much moving around.

A newer variant of the log roll, called the "Zuma maneuver", is probably at least as good as the Log roll. See following.

Supplemental material Movie of the log roll exercise

Link to Youtube videos.

Variant maneuvers: much ado about nothing.

All variant maneuvers are basically subsections of the log-roll. You do not have to read onward (and remember all of the inventor's names, unless you are very interested in this, as the log-roll is the gold standard for all variants of lateral canal BPPV. Avoiding reading onward also avoids a lot of confusion.

Most of these papers violate logic and biomechanics. Most are just incomplete. Some of them are both incomplete and are supported by misleading figures that misrepresent the anatomy of the inner ear. Fortunately, there is now a BPPV viewer (see resources at the end) and you can check on the illustrations yourself.

Vannucchi maneuver, also called "forced prolonged positioning" : incomplete

Simply sleeping with the "affected" ear up has been reported to cure about 75% of patients (see Vannucchi et al, 1997).  This is the second to last step of the log-roll (position 3 above).

Considering the mechanics of the situation, one would expect that bad-ear up would work only for case where the debris is close to coming out already -- i.e. the geotrophic variant of BPPV. It would not be expected to work for the ageotropic variant of lateral canal BPPV. The biophysical rationale for this maneuver is puzzling. Debris probably does not take all night to sediment. 10 minutes, in theory, is plenty.

Another Vannucchi maneuver:

Another Vannucchi maneuver is the "Vannucchi-Asprella" maneuver (2005). In this maneuver, while supine, the head is rapidly rotated away from the "bad" ear, then moved into sitting, then slowly aligned with the body, and then returned to supine. This maneuver attempts to use inertial force to displace otoconia. In our opinion, inertial force is not likely to move debris as intertial force is less than gravitational force, but perhaps it "breaks loose" otoconia adherent to the canal wall. We advise against this one due to lack of logic and also we think it best to avoid rapid head movements for safety..


Gufoni Maneuver for ageotrophic nystagmus, from Appiani et al, 2005. This is an incomplete maneuver, as while it might convert an ageotrophic to a geotrophic nystagmus, it would not logically get the debris "flushed" out entirely (see text).

Gufoni maneuver variants-- one logical one (geotrophic) and one incomplete one (ageotrophic).

The log-roll has 4 positions, and of course, you can create a lot of variant maneuvers by choosing a few of the 4 and leaving out the rest. When some of the 4 positions are unnecessary, you can save some time this way, so it is mainly relevant to busy clinicians.

Appiani and associates (2001; 2005) reviewed several of the quicker  repositioning maneuvers for lateral canal BPPV. Many of them are named after their inventor -- for example, the "Gufoni" maneuver. They generally involve side-lying for 2 minutes, a turn of the head 45 degrees either up or down, remaining in this position for 2 minutes, and then a return to the upright position. Essentially half of the log-roll. Abrupt movements of the head are suggested in these maneuvers - - but in our opinion, there is unlikely to be a benefit from the abrupt movements (Hain et al, 2005), see section below on rapid head movements.  The process is as follows:

For the geotrophic variant, the Gufoni maneuver is reasonable. The debris is almost into the vestibule anyway, and the "business end" of the logroll for geotrophic is the last two positions, which resemble the geotrophic Gufoni

The ageotrophic Gufoni maneuver is not a complete treatment --  The procedure -- lying down one one side, and then turning the head 45 degrees up is half of a log roll, might convert an ageotrophic to a geotrophic BPPV, but it would not be likely to "fix" the patient. To complete the job, one would need to do the other half of the log-roll (i.e. the geotrophic Gufoni) to finish up. In other words, for debris that is 270 degrees away from the vestibule, you need to go the "whole distance", and this 90 degree maneuver is incomplete.

Casani, Gufoni and associates (2011) took this question on head-on, and concluded that both the logroll (which they call barbeque) and the Gufoni were "valid" ways of treating lateral canal BPPV. They also explicitly recognized the "conversion" between ageotrophic and geotrophic.

Similar results were published by a larger group of italian investigators, including many of the same authors in 2013 (Mandala et al, 2013). They reported between 75-84% success at "1 and 24 hour followup", not separating out geotrophic from ageotrophic. We think a week followup would be more relevant. It is puzzling that the Gufoni for ageotrophic did so well, given that it is just a half-logroll maneuver. And why did sham maneuver do so poorly (10%). We can see why it might do poorly at 1 hour follow-up, as presumably nothing was done, but why so poorly at 24 hour after a nights rest (e.g. the Vannuchi forced prolonged positioning maneuver)

Kim et al (2012) recently published a randomized trial of the Gufoni maneuver for the less common, ageotrophic varient of lateral canal BPPV (Kim et al, 2012), and reported a 73% response rate. While we are pleased that an attempt was made to perform a controlled study, this study contained many serious flaws.

The Kim study, like all others of lateral canal BPPV, has the intrinsic problem of determining the side to treat. In the Kim study, there were judgments made based on intensity of nystagmus in several positions, that went without documentation.  The sham maneuver in the Kim study resembled the active maneuver, but was done on the opposite side.  If the side was incorrect, the Sham maneuver might treat. Interestingly, there was a substantial response to the Sham.    Figure 2 in the paper, which is used to explain the mechanism, does depict either known ear anatomy or even a reasonable depiction of rotation of a cartoon of ear anatomy.

For these maneuvers, one must both know if the nystagmus is geotrophic/ageotrophic as well as know the side of debris (which is not always the clearest). Because the usual assumption is that the "affected" side is the one with stronger nystagmus in geotrophic, and the affected side is the side with weaker nystagmus in ageotrophic, in either case, one starts on the side with weaker nystagmus.  If the nystagmus is geotrophic, one after 2 minutes, then one proceeds towards nose down. If it is ageotrophic, then nose up.   We suuggest for the ageotrophic, continuing on with either the "geotrophic" Gufoni or just the rest of the logroll. Nausea or vomiting are obvious potential issues with these maneuvers that require one to spend 4 minutes in positions that induce severe vertigo.

Of course if you are just not sure which side is affected, you could end up mistaken as to the maneuver and end up doing nothing at all.

The main advantage of these abbreviated procedures is that they may be quicker as they cut out some of the useless steps of the Log-roll. We agree with the logic that the full log-roll is not always necessary for geotrophic BPPV. We do not agree with the logic that the "ageotrophic Gufoni" is a complete treatment. We think a log-roll is the logical treatment for ageotrophic.

Zuma maneuver

The Zuma maneuver is basically a log-roll with different timing and an emphasis on rapid head movement.

Zuma's maneuver for ageotrophic lateral canal, from Ramos et al, 2019

In this maneuver there is a combination of gravity and inertial forces. The logic is to begin with the "bad ear down", done "briskly" presumably to attempt to detach otoconia and then let them sediment to the bottom of the lateral canal (e.g. Yacovino et al, 2021). Then another 90 degree turn of the head is made, again rapidly. Following 3 minutes, a third turn is made (we are now 270 degrees from the start). Instead of going nose-down (as in the log roll), the individual is sat up. We find it difficult to follow the anatomy in this figure from the author's paper on the topic, but still we think that the general idea of doing a log-roll with rapid head movements, and long waits between each position is reasonable enough.


We have roughly recreated the anatomy, perhaps in a little more accurate way, using the BPPV simulator of Dr. Teixido.

The lateral canal is red, and the arrow points to the otoconia.

Zuma 1
Zuma 2
Zuma 3
Zuma 4
Illustrations of the geometry of the Zuma Maneuver, made by the author of this page, with the help of the BPPV geometry simulator. Lateral canal is red.

The geometry of this maneuver can be appreciated, and it all makes reasonable sense till position 4 which seems less optimal than nose down (i.e. the original Log roll position).

Places where improvement in the Zuma maneuver might be possiible:


Rapid head movement (about the yaw axis) as a treatment for lateral canal BPPV

Another variant is to move the head briskly towards the good ear during each step, which might add an inertial acceleration component to the repositioning process (Lempert and Tiel-Wielck, 1994). However, theory suggests that inertia contributes little to the movement of otoconia (Hain et al, 2005). Still this is part of some maneuvers (e.g. see the Zuma variant).

Brisk turns adds risk to the maneuver as it could hurt the treated person's neck as well as, in theory at least, dissect a vertebral or carotid in the same way that forceful chiropractic manipulations can sometimes induce stroke. Actually, we encountered one patient who dissected their vertebral from swimming, and another who dissected their carotid artery from VEMP testing with head turning (not the method we favor). . Brisk head movements may also increase risk of retinal detachment. Vibration might be a little safer -- but no studies so far. In spite of all of these thoughts about risk, practically we have never encountered anyone with a dissection or a retinal detachment related to a physical therapy maneuver of this nature.

Head-shaking as a treatment for lateral canal BPPV.

Several authors have suggested that rapid horizontal head-shaking can resolve lateral canal BPPV (Oh et al, 2009; Vanucchi et al, 1997; Kim et al, 2012). This is somewhat plausible considering that the lateral canal is normally tilted so that debris would tend to roll out of it, and by shaking things up, this might be encouraged. 

Kim et al (2012) described the maneuver as follows: " For head-shaking maneuver,15 patients were brought into a sitting position. After pitching the head forward by approximately 30°, we moved the head sideways in a sinusoidal fashion at an approximate rate of 3 Hz for 15 seconds". They do not mention the head excursion (presumably small -- for convenience lets choose 57.2958 degrees = 1 radian). They reported similar rates of response to head-shaking (63%) to the Gufoni (73.1%), both of which were much better than the sham maneuver (34.7%).

In theory, the tangential acceleration of the ear (r=6cm) at 3 hz is r*w (w=angular velocity), which is only about a tenth of a gravity. The problem is the small radius, r. The radial acceleration, which is r*w**2, is higher because it depends on the angular velocity term squared.

There are some interesting issues brought up here -- in our clinic, we routinely use head-shaking as a diagnostic maneuver, prior to positional testing, and prior to any physical treatments. We also use the HIT test, which incoporates a high acceleration head movement. Could we be treating patients by accident ?

See also the comments above about the dangers of brisk head shaking.

Testing for lateral canal BPPV, if it doesn't get better with maneuvers:

Neuroradiological investigation may be warranted in persons who fail to improve after these maneuvers as nystagmus similar to lateral canal BPPV can occur in persons with cerebellar lesions.  We have encountered similar nystagmus in persons with cerebellar lesions. Generally speaking however, when a cerebellar lesion creates positional nystagmus, neurological examination also reveals cerebellar signs other than nystagmus.

Recurrence of lateral canal BPPV

Little is known about recurrence of lateral canal BPPV. Sakaida and others (2003) reported that lateral canal BPPV recurs more frequently than posterior canal BPPV (by about a factor of two). Their report was based on a total of only 19 patients with lateral canal BPPV.  We find this dubious, as the geometry of the lateral canal makes it unlikely to develop lateral canal BPPV in the first place (LC bppv is much less common than PC BPPV), as well as unlikely for a patient to repeat the same process.

Treatment of variant lateral canal BPPV

Heavy cupula: (i.e. ageotrophic, non-fatiguing)

Currently it is generally felt that this is a poor prognosis variant of lateral canal BPPV (although not all agree -- e.g. Bisdorff and Debatisse, 2001), and that it is characterized by ageotrophic nystagmus. As ageotrophic nystagmus could theoretically be cause by debris that is either stuck or attached, this means that there could be two explanations for ageotrophic BPPV, as well as of diverse alternatives (e.g. central nystagmus, ? utricular disturbance, neck nystagmus)..

Biomechanical reasoning (no plunger effect of debris) would suggest that strong ageotrophic nystagmus would necessarily be due to loose debris, and thus that lateral canal cupulolithiasis should not be a big problem because either it is treatable or it is weak (Hain et al, 2005).

For the situation where debris is stuck to the cupula, it may not be easily treated by physical maneuvers aimed at dislodging it (called "liberation of debris by some). Debris could be stuck to either side of the cupula, leading to some uncertainty about which is the best way to treat it. Also, this pattern of nystagmus may derive from central disturbances.

Our approach is to initially try the usual treatments for lateral canal BPPV, possibly with the addition of mastoid vibration. If this fails, we will recommend a variant Brandt-Daroff exercise as tried above. Generally anti-emetic and anti-nausea treatment is necessary when treating lateral canal cupulolithiasis.

Light cupula: (geotrophic, non-fatiguing)

This one is a little hard to discuss as there is actually no method of telling a "light cupula" geotrophic DCPN, from some other non-BPPV positional mechanism such as migraine or a cerebellar disturbance. So all literature about the "light cupula", is usually literature about a conjectured entity. Perhaps it doesn't exist. There are notable exceptions involving temporary changes to blood chemistry -- notably drinking lots of alcohol, and drinking D20, which is heavy water (Money et al, 1965).. Both of these are temporary and just don't apply to chronic clinical situations.

It has been proposed that nonfatiguing geotrophic DCPN might be due to a "light cupula". In other words, a cupula lighter than endolymph which is essentially that of just water (e.g. Imai et al, 2015). This is a bit hard to comprehend as there is just not much material in the body other than fat and air that is lighter than water, and neither one of these is likely to be found in the inner ear. However, as noted above, if protein were high in the endolymph, this might increase its density, and make the cupula relatively lighter than endolymph. Protein does vary considerably among different types of fluids in the inner ear with blood having the most (4238 mg/dl), vs perilymph (178-242), and endolymph (38) (Nyberg et al, 2019).

Kim and Hong (2018) reported that the canalith repositioning maneuver (modified) does not work for this clinical picture, and in fact, " had no therapeutic benefit for a persistent geotropic DCPN and suggest that the pathophysiology of persistent geotropic DCPN is less likely to be a light debris attached to the cupula."

It would seem possible that there might be a very few moving particles in the lateral canal that just move so slowly that it is mistaken for cupulolithiasis. One would think that this would respond to the log roll.

If indeed nothing works, this would suggest that this clinical phenomenon may not be due to otoconia at all. It would not be hard to come up with an alternative explanation involving the otoliths or central processing of utricular signals. When we see this picture in our practice, we will make an attempt to treat with the log roll, and after that we often consider more workup and perhaps treatment for migraine.

Where are Lateral Canal BPPV Evaluations and Treatments Done?

The Vestibular Disorders Association (VEDA) maintains a large and comprehensive list of providers who have indicated a proficiency in treating BPPV. Please contact them to find a local treating provider.

Our own practice is located in Chicago Illinois. We usually combine a visit to a CDH physician with testing for alternatives and then a visit to the physical therapist or a treatment by the physician. Chicago Dizziness and Balance, 645 N. Michigan, Suite 410, Chicago 60611


  1. BPPV viewer (downloadable model of BPPV geometry)



Published literature referred to above: