Compensatory Ocular Torsion

Jampel and Shi¹ are to be congratulated for their recent article in which they used sophisticated digital imaging to confirm their previously published belief that there is no compensatory ocular countertorsion of the eye after head tilt. To me, the methodology seems sound and, at first glance, the hypothesis appears convincing. However, after further reflection, there is something unsettling about accepting their conclusions.

All of the numerous other investigators cited by Jampel and Shi, ¹ myself included, ² consistently reported the presence of a partial compensatory ocular countertorsion after head tilt. I think Jampel and Shi's article falls short in the discussion of why their conclusions differ from those reported by every other investigator. They attempt to address this difference by suggesting that the torsiqo geometric artifact that occurs when angles on a curved surface like the globe are viewed from an eccentric angle. This artifact has previously been described and depicted graphically in detail.² However, false torsion cannot explain the torsion found in many prior studies. Some investigators carefully controlled for false torsion yet still found partial compensatory ocular countertorsion after head tilt.² Similarly, false torsion cannot explain the torsion detected by assessing the change in astigmatic axis after head tilt.²

However, I think the most convincing study documenting the existence of compensatory ocular countertorsion involved a simple afterimage experiment.² With the same afterimage strobe that is commonly used to test for anomalous retinal correspondence in most orthoptic and pediatric eye clinics, a vertical afterimage is created on the retina of 1 eye of the subject. Then a Maddox rod in a trial frame is placed before the eye, and the subject looks toward a light, thus enabling perception of the line created by the Maddox rod. The Maddox rod is rotated so the line it creates is perceived as being exactly parallel to or superimposed on the afterimage. The subject's head is then tilted 30° to 40°. The subject will invariably see the 2 lines (the afterimage and the 1 created by the Maddox rod) rotate with the head. However, the line created by the afterimage will appear to lag behind the Maddox rod, which indicates that the eye did not rotate as far as the head did. By rotating the Maddox rod until the lines are again superimposed, the amount of compensatory ocular countertorsion can be measured and is typically 5° to 10°. I encourage all readers to try this simple experiment on themselves.

Jampel and Shi's¹ findings must be reconciled with these observations. In light of them, yet because of the excellent controls in their experiments, one must conclude that compensatory ocular countertorsion occurs in some situations and not in others. Although it is merely speculation on my part, several possible explanations come to mind. Misslisch et al³ showed that convergence damps ocular torsional movements. In Jampel and Shi's experiments, the fixation target was close to the subject. The convergence that would have occurred in Jampel and Shi's study possibly damped the compensatory ocular countertorsion that obviously occurs in other situations. Also, as I understand Jampel and Shi's experiment, the visual environment tilted with the subject. Perhaps this visual clue had a stabilizing effect that prevented torsion, and no such stabilization occurs when the eye is tilting in reference to the visual environment.

I hope that before Drs Jampel and Shi reply to this letter they try the afterimage experiment on themselves. They will then be able to witness firsthand the compensatory ocular countertorsion that most investigators believe occurs. They can then bring their experience and wisdom on this complex subject to help shed light on why the different experiments resulted in different findings.

Supported by an unrestricted grant from Research to Prevent Blindness Inc, New York, NY, and the Wisconsin Lions Foundation, Stevens Point, Wis, to the Department of Ophthalmology and Visual Sciences, University of Wisconsin, Madison.

Corresponding author: Burton J. Kushner, MD, 2870 University Ave, Suite 206, Madison, WI 53705 (e-mail: bkushner@facstaff.wisc.edu).