Visual–vestibular integration during self-motion perception in younger and older adults.

Abstract

Younger adults integrate visual and vestibular cues to self-motion in a manner consistent with optimal integration; however, little is currently known about whether this process changes with older age. Our objective was to determine whether older adults, like younger adults, display evidence of optimal visual-vestibular integration, including reductions in bimodal variance (Visual + Vestibular) compared with unimodal variance (visual or vestibular alone), and reliability-based cue weighting. We used a motion simulator and a head-mounted display to introduce a 2-interval forced-choice heading estimation task. Older (65+ years) and younger adults (18-35 years) judged which of two movements was more rightward. Movements consisted of vestibular cues (passive movement in darkness), visual cues (optic flow), or both cues combined. The combined condition contained either congruent cues or incongruent cues (either a subtle 5° or larger 20° conflict). Results demonstrated that older adults had less reliable visual heading estimates than younger adults but comparable vestibular heading estimates. During combined, congruent conditions, both age groups exhibited reductions in combined variance, consistent with predicted optimal integration. During subtle cue conflicts, only younger adults exhibited combined variance consistent with predicted optimal integration, but both age groups displayed reliability-based cue weighting. During larger spatial conflicts, neither group demonstrated optimal reductions in variance. Younger adults displayed reliability-based cue weighting but older adults' heading estimates were biased toward the less reliable visual estimate. Older adults' tendency to incorporate spatially conflicting and unreliable visual cues into their self-motion percept may affect their performance on mobility-related tasks like walking and driving. (PsycINFO Database Record

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 49 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.