Square Wave Jerks (SWJ)

Timothy C. Hain, MD • Page last modified: August 1, 2022


Definition of SWJ

Square-Wave Jerks (SWJ) are inappropriate saccades that take the eye off the target, followed by a nearly normal intersaccadic interval (approximately 200 msec), and then a corrective saccade that brings the eye back to the target (Leigh and Zee, 1983). Multiple sources have been suggested as generators for square-wave jerks including the cerebral hemisphere (Sharpe et al 1982), the cerebellum (Alpert et al 1975; Zee et al 1976; Dale et al 1978), and superior colliculus (Hikosaka and Wurtz, 1983).

The figures below illustrate the ENG appearance of SWJ.

Square wave jerks in person with PSP (progressive supranuclear palsy). Recording method was EOG. The larger deviations are saccades in the saccade test. The small superimposed block-like deviations are the SWJ.

Video of abnormally frequent SWJ.

Macro Square Wave Jerks and oscillation in patient with a cerebellar degeneration due to breast cancer. Note how the size of the SWJ increases markedly in the dark. There is also an upbeating nystagmus in the dark. This is a clearly excessive # of square wave jerks. A case of MSO can be found by following the link. See the video link below for what this looks like.

Video of patient with Macro Square Wave JerksVideo of Macro Square Wave Jerks

Double pulses

double pulses

Double pulses is the jargon given to the small spiky eye movements seen above (red). These are brief back-back saccades. They are not associated with a blink as the vertical trace below it shows no deviation. Double pulses are attributed to brainstem disturbances, but practically they have no significance other than this.

All SWJ are horizontal - -there are no vertical or torsional SWJ. See the page on MSO for illustration of this general rule.

Someone who makes large vertical SWJ may be simulating illness.

Disconjugate SWJ.

There are a number of potential causes for a SWJ or a part of a SWJ to look different in one eye than the other.

A new name for one of these phenomena -- dyscongugate "pulses", is "pinball" intrusions -- meaning that when a SWJ is made, the eyes do something differently (i.e. are disconjgate), and also that they alternate according to direction. Lemos et al, 2017, reported these in a case of SCA3. It would seem to us that this is not a specific finding of SCA3, but rather something observed by chance in a cerebellar disturbance. One would think that this might represent a combination of a vergence effort and a disturbance in the saccade "step" generator, and if one seeks these things (which requires hardware that monitors both eyes with high resolution in both eyes), one will find them. Generally speaking, binocular eye movement recordings are not made because they cost at least twice as much as monocular recordings, and add very little utility.

Causes of Square Wave Jerks

The table below shows the commonly reported clinical association -- in essence, nearly anything involving the brain.

  1. Cerebral lesions
  2. Cerebellar disorders
  3. Basal ganglia disorders
  4. Catecholamine depletion
  5. Strabismus
  6. Normal aging

As square-wave jerks are universally found in normal subjects, the main criteria for abnormality is frequency. There are two factors that can affect frequency: age and fixation. Increasing age is associated with increasing frequency. Herishsanu and Sharpe (Herishsanu and Sharpe, 1981) reported a mean frequency of 4.7/min in young and 27/min in elderly. Another factor influencing frequency is the state of fixation. Shallo-Hoffmann and associates ( Shallo-Hoffman et al, 1989) reported that, for normal young subjects, the mean frequency was 4.4/min when recorded in light with visual fixation, 8.5/min when recorded in dark without visual fixation, and 5.4/min when recorded with eyes closed.

SWJ are reported to be more frequent in essential tremor (Gitchel et al, 2013). We have observed this in a single patient with severe essential tremor, as well as retinal disease from diabetes. We are not sure if this is generally the case.


Clinical implications

SWJ are universal (thus SWJ is nonspecific), as well as not present in everyone with brain problems (thus SWJ is insensitive too). The clinical utility of square-wave jerks, which is obviously minor at best, is to point towards the possibility of a central disorder. Because of SWJ are found in everyone, the clinician must be able to judge when SWJ are normal vs. not, from their characteristics.

In young normal persons, square-wave jerks occur infrequently. Accordingly, when frequent SWJ are found in a young patient (more than 1/sec), this should bring up the question of a cerebellar disorder. Macro-square wave oscillations -- MSO -- large and excessive SWJ are an extreme example of this pattern. Follow the link to see a case.

In the elderly, square-wave jerks are common and are rarely of significance. However, in certain conditions such as progressive supranuclear palsy, the diagnosis cannot be made without finding frequent square-wave jerks (Troost and Daroff, 1977). The illustration above is from a patient with PSP.

Technical Issues

No commercial ENG system provides one with any help at all regarding interpretation - -one must use ones clinical judgment.

SWJ may be an artifact of eye movement recording algorithms -- some clinical EOG systems convert blinks into straight lines that appear to be SWJ. It is difficult to know about this unless you turn off the blink-rejection algorithm. Some systems do not allow you to do this -- a huge problem.