Effect of the Stimulus Duration on the Adaptation of the Optokinetic Afternystagmus

Abstract

Observing a rotating visual pattern covering a large portion of the visual field induces optokinetic nystagmus (OKN). If the lights are suddenly switched off, optokinetic afternystagmus (OKAN) occurs. OKAN is hypothesized to originate in the velocity storage mechanism (VSM), a central processing network involved in multi-sensory integration. During a sustained visual rotation, the VSM builds up a velocity signal. After the lights are turned off, the VSM discharges slowly, with OKAN as the neurophysiological correlate. It has been reported that the initial afternystagmus in the direction of the preceding stimulus (OKAN-I) can be followed by a reversed one (OKAN-II), which increases with stimulus duration up to 15 min. In 11 healthy adults, we investigated OKAN following optokinetic stimulus lasting 30 s, 3-, 5-, and 10-min. Analysis of slow-phase cumulative eye position and velocity found OKAN-II in only 5/11 participants. Those participants presented it in over 70% of their trials with longer durations, but only in 10% of their 30 s trials. While this confirms that OKAN-II manifests predominantly after sustained stimuli, it suggests that its occurrence is subject-specific. We also did not observe further increases with stimulus duration. Conversely, OKAN-II onset occurred later as stimulus duration increased (p = 0.02), while OKAN-II occurrence and peak velocity did not differ between the three longest stimuli. Previous studies on OKAN-I, used negative saturation models to account for OKAN-II. As these approaches have no foundation in the OKAN-II literature, we evaluated if a simplified version of a rigorous model of OKAN adaptation could be used in humans. Slow-phase velocity following the trials with 3-, 5-, and 10-min stimuli was fitted with a sum of two decreasing exponential functions with opposite signs (one for OKAN-I and one for OKAN-II). The model assumes separate mechanisms for OKAN-I, representing VSM discharge, and OKAN-II, described as a slower adaptation phenomenon. Although the fit was qualitatively imperfect, this is not surprising given the limited reliability of OKAN in humans. The estimated adaptation time constant seems comparable to the one describing the reversal of the vestibulo-ocular reflex during sustained rotation, suggesting a possible shared adaptive mechanism.