Timothy C. Hain, MD • Page last modified: May 2, 2023
Supplemental material : Video of latent nystagmus. Another video of Latent nystagmus courtesy of Dr. Dario Yacovino (Argentina).
Latent nystagmus (LN) is a variant of congenital nystagmus although some authors prefer to reserve the term congenital nystagmus for other variants. It is mainly encountered in persons with congenital strabismus (specifically esotropia) and amblyopia. In the most common form, the nystagmus appears only when one eye is covered. When both eyes are viewing, no nystagmus is seen. This is the reason why it is called "latent" nystagmus. It can be recognized because the eyes always move with their slow-phase towards the nose for the viewing eye. In other words, the direction of the nystagmus in both eyes changes with the viewing eye.
Most persons with LN will say (if asked) that they see better out of one eye, they may realize that they have no "stereo" meaning that they don't see in 3D, and will often say if asked that they have been told that they have a "lazy eye". So this isn't that hard to diagnose.
The most common variety can be recognized because the eyes nearly always move with their slow-phase towards the nose for the viewing eye. In other words, the direction of the nystagmus in both eyes changes with the viewing eye. The figure below shows a right beating nystagmus, associated with left eye viewing. Recording method is infrared. Notice that the slow phases in this particular patient are not constant velocity but rather speed up as they get closer to the center. This is a common feature of congenital nystagmus, althought not always seen in latent.
Below is LN recorded with a VENG system. There is RB with right eye viewing, and LB with left eye viewing.
![]() |
![]() |
REV SN | LEV SN |
Here is smooth pursuit, that shows the characteristic "back-up" saccades when the pursuit is "riding" the slow phase. This varies according to the viewing eye.
![]() |
![]() |
REV Pursuit -- note "backup saccade" for pursuit going to the L | LEV Pursuit. Note "backup saccades" for pursuit going to the R. |
OPK asymmetries are remarkable in LN. This can be seen at the bedside with a simple hand-held drum. It is not subtle.
![]() |
![]() |
REV OKN -- Good OKN when it rides the nystagmus, nothing in other direction. | LEV OKN, everything is flipped. |
Usually saccades are not much affected, and the vertical system is not affected either.
Latent nystagmus can be very confusing to people attempting to do an ENG. The eyes may jump left or right seemingly at random, and even worse, the unfavored eye commonly deviates to one side depending on which eye is viewing. The author has encountered interesting situations where patients with latent nystagmus have gone unrecognized by otherwise quite competent neurotologists and neuro-ophthalmologists. This is rather silly as all it really takes is thinking of the diagnosis in someone with eyes that are not quite aligned, and then checking to see which way the eyes jump when one eye or the other is covered.
In a person with a strong latent nystagmus, it can be difficult to make much of anything out of the ENG. It is also a source of potentially embarassing blunders.The author has encountered patients who were misdiagnosed by ENT doctors as having more serious conditions (such as a perilymph fistula), because the otherwise quite expert examiner simply did not recognize a strong latent nystagmus.
Usually, if the person doing the ENG did not check for nystagmus with each eye viewing separately, the most reliable method of detecting LN is to observe "backup saccades" during pursuit. So in other words, you can figure out the person has LN after the test, even if the audiologist or technician didn't know about or think of latent nystagmus.
There are a few variants of latent nystagmus.
The most common variant of latent nystagmus is difficult to see with both eyes viewing, and becomes apparent only when one or the other eye is blocked. This is just called "latent nystagmus". With a little practice, persons with latent nystagmus can make their eyes jump to the left or right on command, by looking out of one or the other or both eyes. This was first shown by Van Vliet with his "pseudoscope" (1973), and later by Dell'Osso in a patient with a false eye (Dell'Osso, et al. 1987). We routinely observe this with our video-frenzel goggles -- just ask people to "look" out of one or the other eye, in complete darkness.
Manifest latent nystagmus is latent nystagmus that can be seen even with both eyes viewing. MLN can be acquired, presumably due to a change in the eye that one habitually views. (Dell'Osso et al. 1979).This is unusual compared to regular LN.
In voluntary latent nystagmus, people purposefully view out of either eye, and thus make their eyes jump in either direction at will. This is usually encountered in malingerers trying to convince their doctor to sign disabilty papers or get them out of military service. Obviously, common only in situations where there is something to gain by pretending to be sick.
Dissociated vertical deviation or DVD is a vertical strabismus characterized by a slow upward rotation of one eye without movement of the other. We don't see this very much.
Often persons with latent nystagmus develop a torsional deviation to their eye when they look away from center, which has to be corrected when they return fixation to center. This can be a way of spotting latent nystagmus using video frenzel goggles, during the saccade test. This is common.
Some patients have torsional LN -- their eyes twist in opposite directions depending on the eye that is viewing. Generally they are not very concerned by this, although of course, it must reduce their visual acuity.
A video of torsional CN is shown here. Movie of jerk type torsional Congenital nystagmus in light during fixation (19 meg).
LN is a central nystagmus. The eyes jump according to the intent to view. A person with LN can control the direction of eye jumping (with some practice) through an act of will (van Vliet, 1972). Van Vliet (1972) used a device called a "pseudoscope", where mirrors were used to fool people into thinking that they were seeing out of one eye, when actually they were seeing out of the other. Van Vliet observed that the nystagmus direction went with the intent to view, not the actual eye viewing. This can easily be confirmed at the bedside with the video Frenzel goggles. Just ask the eye to "look" out of one eye in the dark. It helps to hold their finger in front of that eye -- it will be easily seen that the nystagmus varies (in the dark) according to the "viewing" eye -- and it doesn't actually matter that there is nothing to see.
LN appears to be due to growing up without both eyes viewing. There is atrophy of neural pathways that support binocular viewing. People lose the ability to process and fuse input from both eyes together. Often one eye develops a reduction in visual acuity that cannot be corrected with glasses -- amblyopia.
The amblyopia can sometimes be prevented by forcing children to view out of their less favored eye, using patching.
LN is likely caused by a "dying back" of central neurons, due to lack of binocular viewing in infancy. As the neurons are no longer there, LN is not "curable".
Gabapentin is often useful in reducing the speed of congenital nystagmus. Gabapentin increases an inhitory neurotransmitter used in oculomotor function. We nearly always offer the option of taking gabapentin to patients with CN in our clinic setting in Chicago. It seems likely that pregabalin would work too, but we have not tried it as yet. Pregabilin costs more than gabapentin, but is easier to dose.
Recently, it has also been reported that Memantine (an agent which acts on asparate, glutamate and dopamine) is also helpful (Mclean et al, 2007). Glutamate is a major excitatory neurotransmitter. Memantine was used in doses of 40 mg -- greater than is commonly prescribed for other uses. We have never found a patient who responded to memantine, but we have not tried it often.
Surgical treatment is also sometimes offered -- mainly involving moving the eye so that the "null" is located more centrally. We have not found this to be very helpful, and think in general that it is a bad idea.