Beyond the Normal vHIT: Multidomain Vestibular Phenotyping in Patients with Sustained–Impulsive Discordance

This is the authors' abstract. We don't add key points for this paper.

Background/Objectives: Preserved or relatively preserved impulsive vestibulo-ocular responses do not necessarily explain persistent dizziness, motion sensitivity, or incomplete compensation. We investigated whether patients selected for sustained–impulsive discordance show reproducible multidomain vestibular heterogeneity when objective instrument-derived measurements are organized into structured functional domains. Methods: This single-center registry-based observational study analyzed VSD registry data collected from 1 March 2026 to 18 September 2026. Forty-five registry-confirmed/prudent/overlap cases formed the descriptive cohort and forty had usable fast-domain data without a frank dominant impulsive deficit. Before final dataset locking, all records initially flagged as incomplete or non-adjudicable underwent source-document reconciliation against the original instrument-generated reports. This recovered quantitative measurements that had been acquired during the original examinations but had not been fully transcribed into the analytical registry; no new SVIN, OKN, or rotational measurements were generated for phenotype assignment. All 40 fast-eligible patients consequently received final consensus classification on the visual–vestibular and immediate vibration-evoked/SVIN axes. Results: Final consensus classified 37/40 (92.5%) patients as visual–vestibular abnormal and 33/40 (82.5%) as abnormal for immediate vibration-evoked/SVIN response. Cross-classification yielded three observed configurations: 30/40 (75.0%) visual-abnormal/SVIN-evoked, 7/40 (17.5%) visual-abnormal/SVIN-negative or non-phase-locked, and 3/40 (7.5%) visual-preserved/no focal abnormality with SVIN-evoked. Pre-consensus blinded inter-rater agreement was 97.5% for visual–vestibular status (Cohen’s κ = 0.844) and 97.4% for immediate SVIN among 39 paired adjudicable ratings (κ = 0.918). HST was binary-adjudicable in 24/40, with 5/24 (20.8%) positive and 19/24 (79.2%) negative/minimal. VEMP data were adjudicable in 22/40; 12 showed abnormal/dissociated and 10 preserved/symmetric profiles. Conclusions: Preserved or relatively preserved impulsive vestibulo-ocular function did not define a homogeneous vestibular phenotype within this deliberately enriched registry. Complete-cohort multidomain adjudication exposed dissociations among visual–vestibular behavior, vibration-driven responses, head-shaking after-responses, otolithic modifiers, and sustained/post-rotatory dynamics. From a personalized-medicine perspective, this patient-specific functional profiling provides a basis for prospective stratification of domain-targeted diagnostic and rehabilitation strategies, while not establishing a new diagnostic entity or treatment algorithm.