Pooled reference values for the video head impulse test in healthy adults: an observational cross-sectional study

Abstract Background The video head impulse test (vHIT) provides an objective assessment of semicircular canal function through measurement of vestibulo-ocular reflex (VOR) gain and asymmetry. Reference data from healthy adults are important for distinguishing physiological variation from vestibular dysfunction. The need for age-partitioned reference intervals should be supported by evidence of meaningful age-related differences. Aim To establish preliminary pooled reference values for VOR gain and asymmetry ratio (AR) across the lateral, RALP, and LARP semicircular canal planes in healthy adults and to examine the association between age and vHIT parameters. Methods This observational cross-sectional study included 200 healthy adults (18–85 years; observed range). Participants were recruited from hospital staff and community volunteers between 5 February and 15 September 2025. Of 250 individuals screened, 200 were eligible and enrolled and 50 were excluded. Participants underwent otoscopy, pure-tone audiometry, impedance tympanometry, otoacoustic emission testing, and general balance and coordination assessment (Fukuda stepping, sharpened Romberg, and finger–nose pointing tests). VOR gain was measured for six canals using the BalanceEye 3.2-vHIT system (Cyclops Medtech, Bengaluru, India); asymmetry was calculated for the lateral, RALP (RA–LP), and LARP (LA–RP) planes. Age-group comparisons, age correlations, and paired canal comparisons were exploratory. Benjamini–Hochberg FDR correction was applied across 21 tests. Pooled 2.5th and 97.5th percentile limits were estimated across all 200 participants using 5,000 bootstrap replicates. Only 11 participants (5.5%) were aged ≥ 70 years and 3 (1.5%) were aged ≥ 75 years. Results Mean VOR gains were 0.91 ± 0.13 (LL), 1.07 ± 0.13 (RL), 0.92 ± 0.15 (LA), 0.90 ± 0.14 (RP), 1.08 ± 0.14 (RA), and 1.05 ± 0.17 (LP). Mean asymmetry was 9.35 ± 5.51% for the lateral plane, 6.35 ± 5.38% for RALP (RA–LP), and 7.27 ± 5.45% for LARP (LA–RP). No age-group difference or age correlation remained significant after FDR correction. Among paired comparisons, only LL–RL remained significant after correction; LA–RP (adjusted p = 0.170) and RA–LP (adjusted p = 0.079) were not significant. Conclusion This study provides preliminary pooled, device- and population-specific vHIT values in a selected healthy adult sample with observed ages of 18–85 years, although adults ≥ 70 years were sparsely represented ( n = 11; ≥75 years, n = 3). The results do not establish stability of vHIT gain across the oldest decades and do not support separate age-specific reference intervals. The pooled values are not definitive clinical standards and require validation in larger, multicentre cohorts with adequate older-adult representation, test–retest assessment, and cross-platform comparison.