Timothy C. Hain, MD Page last modified: January 4, 2018
Pendular nystagmus refers to the waveform of the involuntary eye movement. In pendular nystagmus the eye moves in sinusoidal trajectory, similar to that of a pendulum. Pendular nystagmus generally does not have a "fast phase" including a saccade, but is composed entirely of slow eye movements.
Pendular nystagmus is generally due to either loss of inhibitory feedback control on a neural circuit, or increased delay in a feedback circuit, that then breaks into spontaneous oscillation. This is a general characteristic of feedback control circuits with delays -- they are vulnerable to increases in gain or increased delay.
Most of the time, pendular nystagmus has a frequency of about 1 hz. This is probably because most pendular nystagmus is due to lesions in the central tegmental tract, which is part of a fairly long loop. A general principal of feedback systems is that longer loops, with longer delays, oscillate slower. If you see a faster nystagmus than 1 hz, it is likely due to a shorter feedback loop, and not part of the central tegmental tract circuitry.
Pendular nystagmus is distinguished from linear slow phase nystagmus such as is seen in vestibular disorders. It is also distinguished from most types of congenital nystagmus, where the waveform commonly has both slow and fast phases.
Pendular nystagmus is a type of oscillation. Oscillations are generally caused by feedback loops. There are two general methods, a delay in a feedback loop, or a "lag" which translates into a delay that changes with frequency. When there is a net positive feedback in a loop, which can be attained in a negative feedback loop with an 180 deg phase lag, the loop can break into oscillation.
The frequency of the oscillation is set by the delay time. For example, a 1 second delay might cause an 1 hz oscillation.
While a fairly typical pendular nystagmus might have a period of 1 second, in the brain, most of the delays are much shorter than 1 sec. For example, it only takes about 200 msec response time for rather complex reactions (i.e. reaction time). Thus to create a nystagmus or tremor that has a frequency of 1 hz, a pure delay of 1 second isn't very likely.
Rather than formulating the problem in terms of neurons sending signals in a loop with a delay, there are also a physiology involved with oscillations called "gap junctions" (Juszcak et al, 2009). Gap junctions may be the mechanism for some of the pendular nystagmus that we encounter such as in oculopalatal myoclonus or in bleeds into the upper brainstem. There are many drugs that affect gap junctions, mainly relatives of quinine, that may have a role in suppressing pendular nystagmus. Little work has been done in this area.
Most frequently, horizontal pendular nystagmus is caused by central lesions involving the central tegmental tract. The concept is that there is a feedback loop controlling eye position or velocity. Lesions add delays, causing instability and nystagmus.
Multiple sclerosis is the most common cause of this sort of nystagmus. Dysmyelinating disorders such as Pelizaeus-Merzbacher is another example (Hobson et al, 2000). Ashoff et al concluded that this type of nystagmus was caused by damage to the cerebellar nuclei, and Gresty et al suggested that the lesion was near the oculomotor nuclei. If this is true, one would think that it could oscillate more quickly than 1 hz (we are dubious).
Gabapentin has been reported as a potential treatment, used in doses of 600-1500 mg/day (Stahl et al, 1995). We ourselves find that gabapentin is generally useful for nystagmus of nearly any type, including congenital nystagmus, suggesting that it works on the "final common pathway".
Spasmus nutans is a transient pendular nystagmus that occurs in children, accompanied by a head tremor and torticollus. The usual age of onset is between 4 and 18 months of age. Sometime patients with SN also have a head tremor.
The nystagmus is generally bilateral (but it can differ in each eye and may even be strictly monocular -- thus it can be dissociated -- see below), and it oscillates in horizontal, torsional, or vertical directions. An instance of spasmus nutans presenting with monocular nystagmus in monozygous twins has been reported.
Spasmus nutans may sometimes be mimicked by tumors of the optic nerve, chiasm, or third ventricle.
|Pendular nystagmus of Spasmus Nutans.|
The nystagmus shown below is from a patient with a pontine hemorrhage who also had oculopalatal myoclonus. The clip below shows her jerk horizontal nystagmus (top), and her pendular vertical nystagmus (bottom). This patient had no horizontal movement in one eye, but both eyes had vertical pendular nystagmus.
Also see movie of her vertical pendular nystagmus.
The most common cause of vertical pendular nystagmus is the vertical nystagmus that frequently appears after pontine hemorrhage. The PPRF is often damaged which makes horizontal saccades difficult, and a vertical pendular nystagmus emerges, possibly related to damage to the central tegmental tract. One wonders why MS lesions of the CTT cause mainly horizontal nystagmus, but pontine bleeds, mainly vertical nystagmus.
Treatment of vertical pendular is difficult. Herishanu and Louzoun reported a single patient who improved after chronic trihexyphenidyl treatment. The doses are relatively massive (1 mg-60 mg), precluding use in most cases because of anticholinergic side effects.
Vertical pendular nystagmus associated with oculopalatal myoclonus syndome. Recording method: infrared video
(c) 2003 Timothy C. Hain, M.D. Educational use is permitted.
The so-called "ocular myoclonus" associated with the oculo-palatal myoclonus syndrome, an example of which shown above, may respond to valproic acid (Lefkowitz and Harpold, 1985). It has also been reported to be successfully treated with INH (200 mg BID to QID, with pyridoxine) combined with converging prisms. In spite of these sporadic reports of success, we have not had much success ourselves. See the OPM page for more details.
Toluene sniffing, usually associated with glue sniffing, also can manifest as a vertical pendular nystagmus, usually in accompaniment with cognitive and cerebellar deficits (Maas et al, 1991).
We have seen one case of vertical pendular nystagmus due to superior canal dehiscence. Here the presumed mechanism is pulsations in CSF pressure stimulating the superior canal via the dehiscenct bone.
References re. vertical pendular nystagmus
Sometimes a horizontal and vertical nystagmus will occur together. They are usually of the same frequency, but their relative phase may different. Depending on the mutual phase, the eyes may take on an oblique direction, form an ellipse. So far, I know of no report of a nystagmus that changes relative phases (causing a pattern where the ellipse was changing). This seems possible however, as in a nystagmus called "windmill nystagmus", the vector of the horizontal nystagmus changes from right-left to up-down and back again, constantly.
Pelizaeus-Merzbacher disease (PMD) is a cause of elliptical nystagmus. It is an inherited disorder, X-linked recessive in children and autosomal dominant in adults. It affects of central white matter, with neonatal, childhood and adult onset forms. Among the leukodystrophies, PMD can be largely distinguished by its binocular pendular nystagmus and head tremor that usually begin at onset. Of the other leukodystrophies, only Cockayne's syndrome has a similar nystagmus. In Cockayne's syndrome, in addition to the findings of PMD, there are also basal ganglia calcifications. In the connatal form of PMD, death usually occurs within a few years of life. In the infantile (classical) onset form, death usually occurs in the second or third decade of life.
- Trobe JD, Sharpe JA, Hirsch DK, Gebarski SS. Nystagmus of Pelizaeus-Merzbacher Disease. Arch Neurol, 48, 1991. 87-91
See the link above for more about torsional nystagmus.
Torsion is difficult to record. One method is to have the person look to the side or up. Then torsion "cross couples" into horizontal or vertical (vertical for lateral gaze, horizontal for vertical gaze). In the recording above, an individual with high-frequency pendular was recorded looking to the side, with resulting cross coupling making it recordable.
Some patients with oculopalatal myoclonus have mainly a torsional nystagmus. We do not know why some develop vertical nystagmus and others torsional nystagmus.
Some congenital nystagmus patients have mainly a torsional pendular nystagmus. One case we encountered has no stereopsis, has a very weak torsional pendular that increases in the light compared to the dark, and seems to experience relatively little impact. In the past, he would move his head, presumably to dampen the nystagmus.
Seesaw nystagmus is a rare binocular disorder characterized by alternating vertical skew deviation and conjugate ocular torsion. Jerk seesaw consists of torsional slow phases in one direction and quick phases in the opposite. In pendular SSN, there are slow, smooth eye oscillations.
- Rambold and others. Seesaw nystagmus associated with involuntary torsional head oscillations. Neurology 1998:51:831-837
DN is a nystagmus where oscillations differ in direction or amplitude between two eyes. The eyes are "dissociated". It has been described in spasmus nutans (see above), congenital nystagmus, and various brainstem disorders, as well as disorders where there has been monocular visual loss from an early age, or acquired later in life. It also occurs when one or more of the muscles controlling an eye are weaker than on the other eye. The nystagmus is generally primarily vertical, has a pendular waveform, is of low frequency and small amplitude. We have mainly seen DN after large cerebellar strokes.
DN is hard to record as it requires binocular cameras.
Despite dissociation, the oscillations of the two eyes are phase locked.
Asymmetrical nystagmus can also be seen when the vergence system is involved in a nystagmus. For example, this can occur after a pontine hemorrhage.
- Bogousslavsky J, Regli F. Convergence and divergence synkinesis. A recovery pattern in benign pontine hematoma. Neuro-ophthalmology. 1984:4:219-225.