Christian Van Nechel ^1,2* Duquesne Ulla ^1,2 Michel Toupet ^2,3 Charlotte Hautefort ^4,5,6,7,8

• ¹ Clinique des Vertiges, Brussels, Belgium
• ² Institut Recherche Oto-Neurologique, Paris, France
• ³ Centre d' Explorations Fonctionnelles Oto-Neurologiques Falguiére, Paris, Île-de-France, France
• ⁴ Assistance Publique Hopitaux De Paris, Paris, Île-de-France, France
• ⁵ Université Paris Cité, Paris, Île-de-France, France
• ⁶ Institut Pasteur, Paris, Île-de-France, France
• ⁷ Fondation Pour l'Audition, Paris, France
• ⁸ Institut de Recherche Oto-Neurologique, Paris, France

While most head movements in daily life are active, most tools used to assess vestibular deficits rely on passive head movements. A single gain value is not sufficient to quantify gaze stabilization efficiency during active movements in vestibular deficit patients. Moreover, during active gaze shifts, anticipatory mechanisms come into play. Our aim was to explore the extent to which previously described compensatory mechanisms are employed in patients with bilateral vestibular loss. We used a Video Head Impulse Test (vHIT) to simultaneously record eye and head movements during a video Active Gaze Shift Test (vAGST). Thirty-eight patients with bilateral vestibular deficits and 61 control subjects were tested. Despite impaired performance on caloric tests and vHIT, most patients exhibited normal gaze stabilization (gain = 0.92 ± 0.1) during active gaze shifts up to a head speed (‘stall speed’) of approximately 140 ± 60°/sec, compared to 280 ± 65°/sec in controls. Our results indicate that BVD patients spontaneously adopt a head speed during active horizontal movements that significantly improves gaze stabilization compared to passive movements. The stall speed correlates with the spontaneous head speed typically adopted by BVD patients and controls in daily activities. As a consequence of the reduction in head speed and corrective saccades, patients also showed an increased delay in gaze stabilization (413 ± 105 ms in BVD patients versus 358 ± 82 ms in controls) at the end of the gaze shift, which might become disabling for certain tasks. Recent model suggests that compensatory eye movements, which stabilize gaze during the counter rotation phase of active gaze shifts, are predictive in nature. vAGST is not designed to provide an etiological diagnosis but rather a functional assessment of the patient’s ability to generate predictive eye movements that compensate for vestibular sensor deficits. Understanding the quality of the patient’s sensory predictions can also shed light on vestibular symptoms, even in cases where no vestibular sensor deficit is detected. This suggest that quality of life and oscillopsia questionnaires should distinguish between predictable and unpredictable movements.