Skip to content

Concurrent validity and agreement of a virtual reality-based Clinical Testclinical test of Sensory Integrationsensory integration and Balancebalance in middle-aged adults.

Sep 2026 · Gait & Posture · Vol 131, pp. 110631 · 0 citations · 26 references
Medicine

Abstract

Background

Sensory integration for postural control relies on dynamic reweighting of visual, somatosensory, and vestibular inputs. The Clinical Test of Sensory Integration and Balance (CTSIB) is widely used, but laboratory-based systems can be costly and less portable. Virtual reality (VR) may provide a scalable alternative, although VR-specific visual features may alter sway behavior and limit interchangeability with traditional protocols.

Purpose

To evaluate the concurrent validity, agreement, and bias of a VR-based CTSIB (VRCTSIB) relative to a traditional force-plate CTSIB across matched sensory contexts.

Methods

Forty adults aged 40-59 years completed 12 conditions (6 CTSIB, 6 VRCTSIB) during one laboratory session. Matched contexts combined surface (firm, foam) and visual state (normal vision, vision deprivation, sway-referenced/visual perturbation). Outcomes were EllipseArea95, total path length, and radial RMS sway. Validity was assessed with Pearson correlations. Agreement was assessed using ICC(2,1) and Bland-Altman analyses, with proportional bias tested by regressing VRCTSIB - CTSIB differences on the method mean.

Results

VRCTSIB and CTSIB were significantly associated across all matched contexts and outcomes (all p < 0.001). Correlations ranged from 0.700 to 0.796 for EllipseArea95, 0.573-0.882 for path length, and 0.542-0.780 for sway. Agreement was condition dependent and lowest in foam eyes-open conditions. Bland-Altman analyses showed context-dependent bias, including notable positive bias in path length during firm eyes-open and foam eyes-open conditions.

Conclusions

VRCTSIB showed evidence of concurrent validity with CTSIB; however, agreement and bias varied across sensory conditions, indicating that the methods should not be considered universally interchangeable without condition-specific calibration or VRCTSIB-specific normative values.

View source

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.